CN102930155A - Method and device for acquiring early-warming parameters of power demands - Google Patents

Abstract

The invention discloses a method and a device for acquiring early warning parameters of power demand. Among them, the method includes: obtaining the data sequence used to generate the early warning index; filtering the data sequence according to the adjustment parameters to obtain the data sequence; calculating the trend index of the data sequence containing the trend item and the period item, and obtaining the early warning index sequence; extracting The benchmark index and the selected index in the early warning index sequence; the correlation calculation is performed according to the time difference analysis model and/or the KL information model to obtain the correlation coefficient between the selected index and the benchmark index, and according to the correlation coefficient. Select the indicators for screening to obtain the leading indicators and consistent indicators; synthesize the leading indicators and consistent indicators to obtain the leading composite index and consistent composite index as early warning parameters. Through the present invention, it is possible to accurately acquire the early warning parameters of electric power demand, thereby accurately formulating reasonable and scientific countermeasures according to short-term periodic fluctuations.

Description

Method and device for obtaining early warning parameters of power demand

technical field

The present invention relates to the field of electric power, in particular, to a method and device for obtaining early warning parameters of electric power demand.

Background technique

The cyclical fluctuation of the economy is an objective phenomenon in economic and social development. It is an objective law that does not depend on human will in the process of economic growth. It is unrealistic to try to forcefully eliminate cyclical fluctuations through various artificial means. The artificially enforced elimination of periodic volatility can also exacerbate volatility. Through the study of the operating law of the economic cycle, we can grasp the fluctuation law of the economic cycle, and then take appropriate measures to reduce the amplitude of the economic cycle fluctuation and prolong the economic fluctuation cycle, so as to achieve the goal of sustainable economic growth.

As a core issue of macroeconomic research, the study of economic cycle has always been valued by the governments of various countries and many economists. With the development of economic theory and the progress of econometric technology, many scholars at home and abroad have begun to use quantitative methods to monitor and predict the economic cycle. The strength of expansion and contraction, etc., so as to formulate scientific and reasonable countermeasures for the government and enterprises according to the characteristics and formation mechanisms of different cycles, slow down the magnitude of cyclical fluctuations, and reduce the damage caused by cyclical fluctuations to economic development.

Economic activities are the driving force of electricity demand, so electricity demand will also fluctuate periodically, but in the field of electricity demand, the means of analyzing periodic fluctuations in electricity demand is to compare the economic growth curve and electricity consumption growth curve for many years, combined with economics According to the classification of my country's economic development stages, experts subjectively judge that the fluctuation cycle of power demand is 9-11 years, and there is no method for studying the periodic fluctuation of short-term power demand.

In order to solve the above problems, forecasting technology can be used to directly predict the future growth rate of electricity demand and analyze its fluctuations. However, forecasting techniques have many flaws:

1. The basic data used has not been adjusted seasonally. The Spring Festival, the number of weekends in a month, the number of holidays, and leap years all have a great impact on the basic data. The fluctuation trend predicted based on this will be distorted;

2. Regardless of using causal relationship models such as regression, per capita electricity consumption, and neural networks, or using time series models such as ARIMA and logistic, reasonable extrapolation is done based on the historical trend of statistical data, which is equivalent to focusing on Figure 1 According to the changing law of medium and long-term trends, it is not possible to quickly make reasonable countermeasures to compound the current economy according to the changing law of long-term trends.

At present, in the field of power demand, forecasting technology is used to obtain short-term periodic fluctuation distortion, and early warning parameters that meet power demand cannot be obtained, resulting in the inability to formulate reasonable measures to deal with periodic fluctuations. No effective solution has been proposed yet. plan.

Contents of the invention

Aiming at the problem in related technologies that using forecasting technology in the field of power demand to obtain short-term periodic fluctuation distortion, it is impossible to obtain early warning parameters that meet power demand, resulting in the inability to formulate reasonable measures to deal with periodic fluctuations. No effective solution has yet been proposed. Therefore, the main purpose of the present invention is to provide a method and device for obtaining early warning parameters of power demand, so as to solve the above problems.

In order to achieve the above object, according to one aspect of the present invention, a method for obtaining early warning parameters of power demand is provided, the method includes: obtaining a data sequence for generating early warning indicators; screening the data sequence according to the adjustment parameters to obtain Data series containing trend items and period items; calculate the trend index of the data series containing trend items and period items, and filter the data series containing trend items and period items according to the trend index to obtain the early warning index sequence, The early warning index sequence is the data whose trend index is increasing in the data sequence; the power generation in the early warning index sequence is extracted as the benchmark index, and the indicators other than the power generation are extracted as the selected indicators; according to the time difference analysis model and/or the K-L information volume model Carry out correlation calculation on the early warning indicators to obtain the correlation coefficient between each selected index and the benchmark index, and screen the selected indicators according to the correlation coefficient to obtain leading indicators and consistent indicators; according to the synthetic index model, the The leading index and the consistent index are synthesized to obtain the leading synthetic index and the consistent synthetic index as early warning parameters.

Further, use the time difference analysis model and/or the KL information model to perform correlation calculations on early warning indicators, and use the correlation coefficient between each selected index and the benchmark index, and filter the selected indicators according to the correlation coefficient, The steps for obtaining the leading index and the consistent index include: obtaining the correlation coefficient _r l between each selected index and the benchmark index according to the following formula:

${\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; =</span>\frac{\underset{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; -</span>\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; -</span>\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; the\; y</span>}}<span\; class="notranslate">\; )</span>}{\sqrt{\underset{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; -</span>\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; -</span>\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; the\; y</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}}<span\; class="notranslate">\; ,</span>$

Among them, l=0,±1,±2,…,±12, Y=(y ₁ ,y ₂ ,…,y _n ) is the benchmark index, X=(x ₁ ,x ₂ ,…,x _n ) is Selected indicators, l is the time difference, n _l is the number of all indicators, t=1, 2, ..., n is the number of months; the value of the time difference is within the first value range and the correlation coefficient r _l is greater than the first The selected index of the threshold is used as the leading index, and the selected index whose time difference value is within the second value range and the correlation coefficient r _l is greater than the second threshold is used as the consistent index.

Further, according to the time difference analysis model and/or the KL information model, the correlation calculation is performed on the early warning indicators, and the correlation coefficient between each selected index and the benchmark index is used, and the selected indicators are screened according to the correlation coefficient, The steps to obtain the leading index and the consistent index include: obtaining the correlation coefficient r _l between each selected index and the benchmark index according to the following formula:

t=1,2,…,n为月份数，l为时差，n_l为所有指标的个数；将时差取值在第三取值范围内且K-L信息量k_l小于第三阈值的标准被选择指标作为先行指标，并将时差取值在第四取值范围内且K-L信息量k_l小于第四阈值的被选择指标作为一致指标。 Among them, l=0,±1,±2,…,±12, Y=(y ₁ ,y ₂ ,…,y _n ) is the benchmark index, X=(x ₁ ,x ₂ ,…,x _n ) is Selected indicators, l is the time difference, n _l is the number of all indicators, t=1,2,...,n is the number of months; the value of the time difference is within the first value range and the correlation coefficient r _l is greater than the first The selected index of the threshold is used as the initial index of the leading index, and the selected index whose time difference value is within the second value range and the correlation coefficient r _l is greater than the second threshold is used as the initial index of the consistent index; for the benchmark index, the leading index The initial index of the index and the initial index of the consistent index are standardized to obtain the standard benchmark index sequence p _t and the sequence q _t of the standard selected index, where the standard selected index includes the standard leading index and the standard consistent index; according to the following formula Obtain the KL information _k between each standard selected index and the standard benchmark index: k _l =∑p _t ln(p _t /q _t+1 ), where l=0,±1,…,±12,

t=1, 2,..., n is the number of months, l is the time difference, n _l is the number of all indicators; the standard that the time difference value is within the third value range and the KL information volume k _l is less than the third threshold is taken Select the index as the leading index, and use the selected index whose time difference value is within the fourth value range and whose KL information volume k _l is less than the fourth threshold as the consistent index.

Further, the step of synthesizing the leading index and the consistent index according to the composite index model to obtain the leading composite index and the consistent composite index as early warning parameters includes: respectively performing symmetrical change processing on the leading index and the consistent index to obtain the leading index symmetry The rate of change C _w,i (t) and the symmetrical rate of change C _z,i (t) of the consistent index, wherein, the leading index is processed symmetrically by the following formula to obtain the symmetrical rate of change C _w,i (t) of the leading index :

其中，

是第i(i=1，2，…，k_w)个先行指标，t=2,3,…,n，k_w为先行指标的个数；通过如下公式对一致指标进行对称变化处理，以获取一致指标对称变化率C_z,i(t)：

其中，

是第i(i=1，2，…，k_z)个一致指标，t＝2,3,…,n为月份数，k_z是一致指标的个数；对先行指标对称变化率C_w,i(t)和一致指标对称变化率C_z,i(t)进行标准化处理和趋势调整之后得到的结果进行合成计算，以获取先行合成指数和一致合成指数。

in,

is the i-th (i=1, 2, ..., k _w ) leading index, t=2, 3,..., n, k _w is the number of leading indexes; the consistent index is changed symmetrically by the following formula, with Obtain the symmetrical rate of change C _z,i (t) of the consistent index:

in,

is the i-th (i=1, 2, ..., k _z ) consistent index, t = 2, 3, ..., n is the number of months, k _z is the number of consistent indicators; the symmetric rate of change of the leading index C _w, The results obtained after standardization and trend adjustment of _i (t) and the consistent index symmetric rate of change C _z,i (t) are combined and calculated to obtain the leading composite index and the consistent composite index.

t=2,3,…,n；采用标准化因子A_w,i和A_z，i分别将先行指标对称变化率C_w,i(t)和一致指标对称变化率C_z,i(t)进行标准化处理，以得到标准化变化率S_w,i(t)和S_z,i(t)，其中，

t＝2,3,…,n；对标准化变化率S_w,i(t)和S_z,i(t)进行平均变化率处理，以获取先行指标的标准化平均变化率V_w(t)和一致指标的标准化平均变化率V_z(t)；根据先行指标的标准化平均变化率V_w(t)和一致指标的标准化平均变化率V_z(t)进行合成计算，以获取先行合成指数I_w(t)和一致合成指数I_z(t)，其中，

$I_z\left(t\right)=I_z(t-1)\&times;\frac{200+V_z\left(t\right)}{200-V_z\left(t\right)},$ 且I_w(1)=100，I_z(1)=100。Further, the results obtained after standardization and trend adjustment of the leading index symmetrical rate of change C _w,i (t) and the consistent index symmetrical rate of change C _z,i (t) are synthetically calculated to obtain the leading synthetic index and the consistent index The steps of synthesizing the index include: obtaining the standardized factors A _w,i and A _z,i by the following formula:

t=2,3,...,n; use the standardization factors A _w,i and A _z,i to carry out the symmetrical change rate C _w,i (t) of the leading index and the symmetrical change rate C _z,i (t) of the consistent index respectively Standardized processing to obtain the standardized rate of change S _w,i (t) and S _z,i (t), where,

t=2,3,...,n; carry out average change rate processing on the standardized change rates S _w,i (t) and S _z,i (t) to obtain the standardized average change rates V _w (t) and The normalized average rate of change V _z (t) of the consistent index; the synthetic calculation is carried out according to the normalized average rate of change V _w (t) of the leading index and the normalized average rate of change V _z (t) of the consistent index to obtain the leading synthetic index I _w (t) and the consensus composite index I _z (t), where,

$I_z\left(t\right)=I_z(t-1)\&times;\frac{200+V_z\left(t\right)}{200-V_z\left(t\right)},$ And I _w (1)=100, I _z (1)=100.

其中，λ_w,i和λ_z,i分别是先行指标和一致指标的第i个指标的权重；通过如下公式获取指标标准化因子F_w： $F_w=\left[\underset{t=2}{\overset{n}{\mathrm{\&Sigma;}}}\right|R_w\left(t\right)|/(n-1)]/\left[\underset{t=2}{\overset{n}{\mathrm{\&Sigma;}}}\right|R_z\left(t\right)|/(n-1)];$ 根据指标标准化因子F_w进行标准化平均变化率处理，以获取先行指标的标准化平均变化率V_w(t)和一致指标的标准化平均变化率V_z(t)，其中，V_w(t)＝R_w(t)/F_w，V_z(t)＝R_z(t)。Further, average rate of change processing is performed on the standardized rate of change S _w,i (t) and S _z,i (t) to obtain the normalized average rate of change V _w (t) of the leading index and the standardized average rate of change of the consistent index The step of V _z (t) includes: using the following formula to process the normalized rate of change S _w,i (t) of the leading index and the standardized rate of change S _z,i (t) of the consistent index respectively, to obtain the leading The average rate of change R _w (t) of indicators and the average rate of change R _z (t) of consistent indicators:

Among them, λw _,i and λz _,i are the weights of the leading indicator and the i-th indicator of the consistent indicator respectively; the indicator standardization factor F _w is obtained by the following formula: $f_w=\left[\underset{t=2}{\overset{\mathrm{no}}{\mathrm{\&Sigma;}}}\right|R_w\left(t\right)|/(\mathrm{no}-1)]/\left[\underset{t=2}{\overset{\mathrm{no}}{\mathrm{\&Sigma;}}}\right|R_z\left(t\right)|/(\mathrm{no}-1)];$ According to the standardization factor F _w of the index, the normalized average rate of change is processed to obtain the normalized average rate of change V _w (t) of the leading index and the normalized average rate of change V _z (t) of the consistent index, where V _w (t) = R _w (t)/F _w , V _z (t) = R _z (t).

Further, after obtaining the data sequence used to generate the early warning indicator, the method also includes: preprocessing the data in the data sequence, the preprocessing includes: filling missing data processing, correcting noise data processing, data smoothing processing, and data normalization processing.

In order to achieve the above object, according to one aspect of the present invention, a device for obtaining early warning parameters of power demand is provided, the device includes: a first obtaining module, used to obtain a data sequence for generating early warning indicators; a first processing module , used to filter the data series according to the adjustment parameters to obtain the data series containing trend items and period items; the first calculation module is used to calculate the trend index of the data series containing trend items and period items, and according to the trend The index filters the data sequence containing the trend item and the period item to obtain the early warning index sequence. The early warning index sequence is the data whose trend index is increasing in the data sequence; the first extraction module is used to extract the power generation in the early warning index sequence as the benchmark index, and extract the index other than the power generation as the selected index; the second calculation module is used to perform correlation calculation on the early warning index according to the time difference analysis model and/or the K-L information model, so as to obtain each selected index The correlation coefficient between the index and the benchmark index, and the selected index is screened according to the correlation coefficient to obtain the leading index and the consistent index; the second processing module is used to synthesize the leading index and the consistent index according to the synthetic index model, To obtain the leading composite index and consistent composite index as early warning parameters.

其中，l=0，±1，±2，…，±12，Y=(y₁,y₂,…,y_n)为基准指标，X=(x₁,x₂,…,x_n)为被选择指标，l为时差，n_l为所有指标的个数，t＝1，2,…,n为月份数；第一子处理模块，用于将时差取值在第一取值范围内且相关性系数r_l大于第一阈值的被选择指标作为先行指标，并将时差取值在第二取值范围内且相关性系数r_l大于第二阈值的被选择指标作为一致指标。Further, the second calculation module includes: a first sub-calculation module, which is used to obtain the correlation coefficient r _l between each selected index and the benchmark index according to the following formula:

Among them, l=0,±1,±2,…,±12, Y=(y ₁ ,y ₂ ,…,y _n ) is the benchmark index, X=(x ₁ ,x ₂ ,…,x _n ) is Selected indicators, l is the time difference, n _l is the number of all indicators, t=1, 2,..., n is the number of months; the first sub-processing module is used to set the time difference within the first value range and The selected index whose correlation coefficient r _l is greater than the first threshold is taken as the leading index, and the selected index whose time difference value is within the second value range and whose correlation coefficient r _l is greater than the second threshold is regarded as the consistent index.

其中，l=0，±1，±2，…，±12，Y=(y₁,y₂,…,y_n)为基准指标，X=(x₁,x₂,…,x_n)为被选择指标，l为时差，n_l为所有指标的个数，t＝1，2，…,n为月份数；第二子处理模块，用于将时差取值在第一取值范围内且相关性系数r_l大于第一阈值的被选择指标作为先行指标的初始指标，并将时差取值在第二取值范围内且相关性系数r_l大于第二阈值的被选择指标作为一致指标的初始指标；第三子处理模块，用于对基准指标、先行指标的初始指标以及一致指标的初始指标进行标准化处理，以获取标准基准指标序列p_t、标准被选择指标的序列q_t，其中，标准被选择指标包括标准先行指标以及标准一致指标；第三子计算模块，用于按如下公式获取每个标准被选择指标和标准基准指标之间的K-L信息量_kl：k_l=∑p_tln(p_t/q_t+1)，其中，l=0,±1，…,±12，

t=1，2,…,n为月份数，l为时差，n_l为所有指标的个数；Further, the second calculation module includes: a second sub-calculation module, which is used to obtain the correlation coefficient r _l between each selected indicator and the benchmark indicator according to the following formula:

Among them, l=0,±1,±2,…,±12, Y=(y ₁ ,y ₂ ,…,y _n ) is the benchmark index, X=(x ₁ ,x ₂ ,…,x _n ) is Selected indicators, l is the time difference, n _l is the number of all indicators, t=1, 2, ..., n is the number of months; the second sub-processing module is used to set the time difference within the first value range and The selected index whose correlation coefficient r _l is greater than the first threshold is used as the initial index of the leading index, and the selected index whose time difference value is within the second value range and the correlation coefficient r _l is greater than the second threshold is used as the consistent index Initial indicators; the third sub-processing module is used to standardize the benchmark indicators, the initial indicators of the leading indicators and the initial indicators of the consistent indicators, so as to obtain the standard benchmark indicator sequence p _t and the standard selected indicator sequence q _t , wherein, Standard selected indicators include standard leading indicators and standard consistent indicators; the third sub-calculation module is used to obtain the KL information _k l between each standard selected indicator and standard benchmark indicator according to the following formula: k _l =∑p _t ln(p _t /q _t+1 ), where, l=0,±1,…,±12,

t=1, 2,..., n is the number of months, l is the time difference, n _l is the number of all indicators;

The fourth sub-processing module is used to take the time difference value within the third value range and the standard selected index whose KL information volume k _l is less than the third threshold as the leading index, and set the time difference value within the fourth value range And the selected index whose KL information amount k _l is less than the fourth threshold is taken as the consistent index.

其中，是第i(i=1，2，…，k_w)个先行指标，t=2,3,…,n为月份数，k_w为先行指标的个数；第五子计算模块，用于通过如下公式对一致指标进行对称变化处理，以获取一致指标对称变化率C_z,i(t)：

其中，

是第i(i=1，2，…，k_z)个一致指标，t=2,3,…,n，k_z是一致指标的个数；第六子处理模块，用于对先行指标对称变化率C_w,i(t)和一致指标对称变化率C_z,i(t)进行标准化处理和趋势调整之后得到的结果进行合成计算，以获取先行合成指数和一致合成指数。Further, the second processing module includes: a fifth sub-processing module, which is used to perform symmetrical change processing on the leading index and the consistent index respectively, so as to obtain the symmetrical rate of change C _w,i (t) of the leading index and the symmetrical rate of change C of the consistent index _{z, i} (t), the fifth sub-processing module includes: a fourth sub-calculation module, which is used to perform symmetrical change processing on the leading index through the following formula, so as to obtain the symmetrical rate of change C _w,i (t) of the leading index:

in, is the i-th (i=1, 2, ..., k _w ) leading indicator, t=2, 3, ..., n is the number of months, and k _w is the number of leading indicators; the fifth sub-calculation module is used to pass The following formula performs symmetrical change processing on the consistent index to obtain the symmetrical change rate C _z,i (t) of the consistent index:

in,

is the i-th (i=1, 2,...,k _z ) consistent index, t=2,3,...,n, k _z is the number of consistent indexes; the sixth sub-processing module is used to be symmetrical to the leading index The results obtained after standardization and trend adjustment of the rate of change C _w,i (t) and the symmetrical rate of change of the consistent index C _z,i (t) are combined and calculated to obtain the leading composite index and the consistent composite index.

t=2,3,…,n；Further, the sixth sub-processing module includes: a sixth sub-calculation module, which is used to obtain the normalization factors A _{w, i} and A _{z, i} through the following formula: t=2,3,...,n; the seventh sub-processing module is used to use normalization factors A _w,i and A _z,i to respectively convert the leading index symmetric rate of change C _w,i (t) and the consistent index symmetric rate of change C _z,i (t) is standardized to obtain the standardized rate of change S _w,i (t) and S _z,i (t), where,

t=2,3,...,n;

The eighth sub-processing module is used to process the average rate of change on the standardized rate of change S _w,i (t) and S _z,i (t), so as to obtain the standardized average rate of change V _w (t) of the leading index and the consistent index The standardized average rate of change V _z (t); the seventh sub-calculation module is used to perform synthetic calculations based on the standardized average rate of change V _w (t) of the leading index and the standardized average rate of change V _z (t) of the consistent index, to Obtain the leading synthetic index I _w (t) and the consistent synthetic index I _z (t), where, $I_z\left(t\right)=I_z(t-1)\&times;\frac{200+V_z\left(t\right)}{200-V_z\left(t\right)},$ And I _w (1)=100, I _z (1)=100.

Further, the eighth sub-processing module includes: a ninth sub-processing module, which is used to respectively convert the standardized rate of change S _w,i (t) of the leading index and the standardized rate of change S _z,i (t) of the consistent index through the following formula Perform average rate of change processing to obtain the average rate of change R _w (t) of leading indicators and the average rate of change R _z (t) of consistent indicators:

$R_w\left(t\right)=\frac{\underset{i=1}{\overset{k_w}{\mathrm{\&Sigma;}}}{S}_{w,i}\left(t\right){\mathrm{\&lambda;}}_{w,i}}{\underset{i=1}{\overset{k_w}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_{w,i}},$ $R_z\left(t\right)=\frac{\underset{i=1}{\overset{k_z}{\mathrm{\&Sigma;}}}{S}_{z,i}\left(t\right){\mathrm{\&lambda;}}_{z,i}}{\underset{i=1}{\overset{k_Z}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_{z,i}},$ Among them, λw _,i and λz _,i are the weights of the leading indicator and the i-th indicator of the consistent indicator respectively; the eighth sub-calculation module is used to obtain the indicator standardization factor F _w through the following formula: $f_w=\left[\underset{t=2}{\overset{\mathrm{no}}{\mathrm{\&Sigma;}}}\right|R_w\left(t\right)|/(\mathrm{no}-1)]/\left[\underset{t=2}{\overset{\mathrm{no}}{\mathrm{\&Sigma;}}}\right|R_z\left(t\right)|/(\mathrm{no}-1)];$ The ninth sub-computing module is used to process the standardized average rate of change according to the index standardization factor _Fw , so as to obtain the standardized average rate of change _Vw (t) of the leading index and the standardized average rate of change _Vz (t) of the consistent index, where , V _w (t)=R _w (t)/F _w , V _z (t)=R _z (t).

Further, after executing the acquisition module, the device further includes: a tenth sub-processing module, which is used to preprocess the data in the data sequence, and the preprocessing includes: filling missing data processing, correcting noise data processing, data smoothing processing, and data Normalized processing.

Through the method and device for obtaining early warning parameters of electric power demand of the present application, after obtaining the trend item and cycle item in the original data sequence, the leading index and the consistent index are obtained by screening and calculating the data sequence, and then the above leading index and the consistent The index is synthesized with the index synthesis model to obtain the early warning parameters, and the periodic fluctuation of power demand is analyzed according to the early warning parameters, which solves the problem of using forecasting technology to obtain short-term periodic fluctuation distortion in the field of power demand in the prior art, and the early warning parameters that meet the power demand cannot be obtained. , so that it is impossible to formulate reasonable measures to deal with periodic fluctuations according to the periodic fluctuations of power, and realize the accurate acquisition of early warning parameters of power demand, so as to accurately formulate reasonable and scientific countermeasures according to short-term periodic fluctuations, thereby slowing down the cycle The magnitude of fluctuations can reduce the damage caused by cyclical fluctuations to the power industry and economic development.

Description of drawings

The accompanying drawings described here are used to provide a further understanding of the present invention and constitute a part of the application. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations to the present invention. In the attached picture:

Fig. 1 is a schematic diagram according to the periodic fluctuation of power demand in the prior art;

Fig. 2 is a schematic structural diagram of a device for obtaining early warning parameters of power demand according to the present invention;

3 is a flow chart of a method for obtaining early warning parameters of power demand according to an embodiment of the present invention;

4 is a detailed flow chart of a method for obtaining early warning parameters of power demand according to an embodiment of the present invention;

FIG. 5 is a schematic diagram of a method for obtaining consistent indicators and leading indicators according to the embodiment shown in FIG. 4; and

FIG. 6 is a schematic diagram of trend adjustment according to the embodiment shown in FIG. 4 .

Detailed ways

It should be noted that, in the case of no conflict, the embodiments in the present application and the features in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and examples.

Fig. 2 is a schematic structural diagram of a device for obtaining early warning parameters of power demand according to the present invention. As shown in Figure 2, the device includes: an acquisition module 10, used to acquire a data sequence for generating early warning indicators; a first processing module 30, used to filter the data sequence according to the adjustment parameters, to obtain the data sequence containing trend items and The data sequence of the periodic item; the first calculation module 50 is used to calculate the trend index of the data sequence containing the trend item and the periodic item, and filter the data sequence containing the trend item and the periodic item according to the trend index to obtain an early warning The index sequence, the early warning index sequence is the data whose trend index is increasing in the data sequence; the first extraction module 70 is used to extract the power generation in the early warning index sequence as the benchmark index, and extract the indexes other than the electricity consumption as the selected index ; The second calculation module 90 is used to calculate the correlation of the early warning indicators according to the time difference analysis model and/or the K-L information model, so as to obtain the correlation coefficient between each selected index and the benchmark index, and according to the correlation coefficient Screen the selected indicators to obtain leading indicators and consistent indicators; the second processing module 110 is used to synthesize the leading indicators and consistent indicators according to the composite index model to obtain the leading composite index and consistent composite index as early warning parameters.

Using the device for obtaining early warning parameters of power demand of the present invention, the data sequence used to generate early warning indicators will be obtained through the acquisition module, and then the first processing module will filter the data sequence and obtain the data sequence containing trend items and period items , then the first calculation module calculates the trend index of the above data sequence, and filters the above data sequence according to the trend index to obtain the early warning index sequence with the trend index in the data sequence growing, and then extracts the early warning index sequence through the first extraction module The power generation of the system is used as the benchmark index, and the indicators other than electricity consumption are extracted as the selected indicators, and then the second calculation module performs correlation calculation on the early warning indicators according to the time difference analysis model and/or the K-L information model, so as to obtain each selected indicator. Select the correlation coefficient between the index and the benchmark index, and screen the selected index according to the correlation coefficient to obtain the leading index and the consistent index, and finally the second processing module synthesizes the leading index and the consistent index according to the synthetic index model, To obtain the leading composite index and consistent composite index as early warning parameters. Through the device for obtaining early warning parameters of electric power demand in the present application, the trend item and period item in the data sequence are obtained first, and then the data sequence is processed to obtain the leading index and the consistent index, and the above leading index and the consistent index are combined into an index synthesis model The early warning parameters are synthesized to solve the problem in the prior art that using forecasting technology to obtain short-term periodic fluctuations in the field of power demand cannot obtain early warning parameters that meet power demand, resulting in the inability to formulate reasonable countermeasures based on periodic power fluctuations , to realize the accurate acquisition of early warning parameters of power demand, so as to accurately formulate reasonable countermeasures according to short-term periodic fluctuations, thereby slowing down the magnitude of periodic fluctuations and reducing the degree of damage caused by periodic fluctuations to the power industry and economic development.

Specifically, in the field of electricity demand, the data sequence used to generate early warning indicators includes four parts: long-term trend items, cycle items caused by periodic laws such as industry expansion-prosperity contraction-recession-re-expansion, spring, summer, autumn and winter or the number of days per month Seasonal items caused by different factors, random items caused by unknown factors. Through the device of the present invention for obtaining the early warning parameters of electric power demand, the seasonal item and the random item in the original data sequence are first eliminated, and the data sequence of the trend item and the cycle item is obtained, and then the quarterly GDP or the monthly industrial added value As a benchmark index, use time difference correlation analysis or K-L information technology to screen several leading indicators and consistent indicators from a large number of other economic indicators, and finally use the index synthesis model to synthesize several leading indicators into leading indexes respectively, and combine several consistent indicators Synthesized into a consensus index, and obtained the early warning parameters, the purpose of judging the future trend of the consensus index by using the advanced nature of the leading index can be realized according to the early warning parameters.

In the above-mentioned embodiments of the present invention, the data sequence is screened by the first processing module 30, and the original data sequence is seasonally adjusted, that is, the influence of seasonal factors and random factors is eliminated, and the long-term trend item (that is, in the above-mentioned embodiment) Based on the trend item) and the short-term cycle item (that is, the cycle item in the above-mentioned embodiment), through index screening and index synthesis, the fluctuation trend of power demand is studied with the boom early warning technology, and the above-mentioned embodiment uses the economic early warning power demand, Select leading indicators and consistent indicators from a large number of economic indicators to judge the periodic fluctuations of power demand, making the judgment results more accurate and making the periodic fluctuations used to formulate reasonable countermeasures more accurate and reasonable.

In the above practical example of the present application, the second calculation module 90 includes: a first sub-calculation module, which is used to obtain the correlation coefficient r _l between each selected index and the benchmark index according to the following formula:

其中，l=0，±1，±2，…，±12，Y＝(y₁,y₂,…,y_n)为基准指标，X=(x₁,x₂,…,x_n)为被选择指标，l为时差，n_l为所有指标的个数；第一子处理模块，用于将时差取值在第一取值范围内且相关性系数r_l大于第一阈值的被选择指标作为先行指标，并将时差取值在第二取值范围内且相关性系数r_l大于第二阈值的被选择指标作为一致指标。

Among them, l=0,±1,±2,…,±12, Y=(y ₁ ,y ₂ ,…,y _n ) is the benchmark index, X=(x ₁ ,x ₂ ,…,x _n ) is Selected indicators, l is the time difference, n _l is the number of all indicators; the first sub-processing module is used to select the selected indicators whose time difference value is within the first value range and the correlation coefficient r _l is greater than the first threshold As the leading index, the selected index whose time difference value is within the second value range and whose correlation coefficient r _l is greater than the second threshold is taken as the consistent index.

Specifically, the benchmark indicators are used as the "benchmark" for screening, and the time difference correlation analysis model is used to initially screen consistent indicators and leading indicators. All the selected indexes except the benchmark indexes lead or lag period l (i.e. the time difference in the above-mentioned embodiment) (l=0, ±1, ±2,..., ±12), the first sub-calculation module is respectively according to the following formula Calculate the correlation coefficient r _l between each selected index and the benchmark index:

$r_l=\frac{\underset{t=1}{\overset{{\mathrm{no}}_l}{\mathrm{\&Sigma;}}}(x_{t+l}-\stackrel{\&OverBar;}{x})({\mathrm{the\; y}}_t-\stackrel{\&OverBar;}{\mathrm{the\; y}})}{\sqrt{\underset{t=1}{\overset{{\mathrm{no}}_l}{\mathrm{\&Sigma;}}}{(x_{t+l}-\stackrel{\&OverBar;}{x})}^2{({\mathrm{the\; y}}_t-\stackrel{\&OverBar;}{\mathrm{the\; y}})}^2}},$ l=0, ±1, ±2, ..., ±12,

和分别为序列X和Y的平均值。n_l是所有指标的数据个数。则最大的时差相关系数被认为反映了被选指标与基准指标的时差相关关系，相应的延迟数l表示超前或滞后期，即使得时差相关性系数r_l最大的延迟数l就是该被选指标与基准指标的超前或滞后期。In the above formula, Y=(y ₁ ,y ₂ ,…,y _n ) is the benchmark index, X=(x ₁ ,x ₂ ,…,x _n ) is the selected index, r is the correlation coefficient, l means leading or lagging Period (instantaneous difference), when l is negative, it means ahead, when it is positive, it means lag, l is called time difference or delay number,

and are the mean values of sequences X and Y, respectively. n _l is the number of data of all indicators. Then the largest time-difference correlation coefficient is considered to reflect the time-difference correlation relationship between the selected index and the benchmark index, and the corresponding delay number l represents the leading or lagging period, that is, the largest delay number _l of the time-difference correlation coefficient r l is the selected index The lead or lag period with respect to the benchmark metric.

According to the above-mentioned embodiment of the present application, the second calculation module 90 may further include: a second sub-calculation module, configured to obtain the correlation coefficient r _l between each selected index and the benchmark index according to the following formula:

其中，l=0，±1，±2，…，±12，Y＝(y₁,y₂,…,y_n)为基准指标，X＝(x₁,x₂,…,x_n)为被选择指标，l为时差，n_l为所有指标的个数；第二子处理模块，用于将时差取值在第一取值范围内且相关性系数r_l大于第一阈值的被选择指标作为先行指标的初始指标，并将时差取值在第二取值范围内且相关性系数r_l大于第二阈值的被选择指标作为一致指标的初始指标；第三子处理模块，用于对基准指标、先行指标的初始指标以及一致指标的初始指标进行标准化处理，以获取标准基准指标序列p_t、标准被选择指标的序列q_t，其中，标准被选择指标包括标准先行指标以及标准一致指标；第三子计算模块，用于按如下公式获取每个标准被选择指标和标准基准指标之间的K-L信息量k_l：

Among them, l=0,±1,±2,…,±12, Y=(y ₁ ,y ₂ ,…,y _n ) is the benchmark index, X=(x ₁ ,x ₂ ,…,x _n ) is Selected indicators, l is the time difference, n _l is the number of all indicators; the second sub-processing module is used to select the selected indicators whose time difference value is within the first value range and the correlation coefficient r _l is greater than the first threshold As the initial index of the leading index, the selected index whose time difference value is within the second value range and the correlation coefficient r _l is greater than the second threshold is used as the initial index of the consistent index; the third sub-processing module is used for benchmarking Indicators, the initial indicators of the leading indicators and the initial indicators of the consistent indicators are standardized to obtain the standard benchmark index sequence p _t and the sequence q _t of the standard selected indicators, where the standard selected indicators include the standard leading indicators and the standard consistent indicators; The third sub-calculation module is used to obtain the KL information amount k _l between each standard selected index and the standard benchmark index according to the following formula:

t＝1，2,…,n，l为时差，n_l为所有指标的个数；第四子处理模块，用于将时差取值在第三取值范围内且K-L信息量k_l小于第三阈值的标准被选择指标作为先行指标，并将时差取值在第四取值范围内且K-L信息量k_l小于第四阈值的被选择指标作为一致指标。其中，第一取值范围可以是小于-3，第二取值范围可以是大于或等于-2且小于或等于2，第一阈值可以是0.7，第二阈值也可以是0.7。其中，t为月份数。k _l =∑p _t ln(p _t /q _t+1 ), where, l=0,±1,…,±12,

t=1, 2,..., n, l is the time difference, n _l is the number of all indicators; the fourth sub-processing module is used to set the time difference value within the third value range and the KL information amount k _l is less than the first The three-threshold standard selects the index as the leading index, and takes the selected index whose time difference value is within the fourth value range and the KL information k _l is less than the fourth threshold as the consistent index. Wherein, the first value range may be less than -3, the second value range may be greater than or equal to -2 and less than or equal to 2, the first threshold may be 0.7, and the second threshold may also be 0.7. Among them, t is the number of months.

Specifically, after the correlation coefficient r _l is obtained through the calculation of the second sub-calculation module, the second sub-processing module uses the selected index whose time difference value is within the first value range and the correlation coefficient r _l is greater than the first threshold as The initial index of the leading index, and the selected index whose time difference value is within the second value range and the correlation coefficient r _l is greater than the second threshold is used as the initial index of the consistent index, and then the third sub-processing module standardizes the benchmark index Processing, the processed standard benchmark index sequence is denoted as p _t :

$p_t={\mathrm{the\; y}}_t/\underset{t=1}{\overset{\mathrm{no}}{\mathrm{\&Sigma;}}}{\mathrm{the\; y}}_t,$ t=1,2,...,n,

The third sub-processing module also standardizes the pre-selected leading indicators and consistent indicators, and the processed standard is recorded as the sequence of selected indicators as q _t :

$q_t=x_t/\underset{t=1}{\overset{\mathrm{no}}{\mathrm{\&Sigma;}}}{x}_t,$ t=1,2,...,n,

Then, the fourth sub-processing module calculates the amount of KL information k _l about the benchmark index after the delay l of each primary index:

k _l =∑p _t ln(p _t /q _t+1 ), l=0,±1,…,±12

Among them, l represents the leading or lagging period. When l is a negative number, it means leading, and when it is positive, it means lagging. _l is called the time difference. In the formula, t represents the month and j represents the year.

其相对应的延迟数l^*就是被选指标最适当的超前或滞后月数（季度）。K-L信息量越接近于0，说明指标x与基准指标y越接近。并将筛选出来的先行指标和一致指标分别记为W⁽³⁾和Z⁽³⁾。More specifically, after calculating 2L+1 KL information amounts, select a minimum value k _l′ from the k _l values as the KL information amount of the selected index x with respect to the benchmark index y, namely

The corresponding delay number l ^* is the most appropriate leading or lagging months (quarters) of the selected indicators. The closer the KL information is to 0, the closer the index x is to the benchmark index y. And the screened leading indicators and consistent indicators are recorded as W ⁽³⁾ and Z ⁽³⁾ respectively.

According to the above-mentioned embodiments of the present application, the second processing module 110 may include: a fifth sub-processing module, configured to perform symmetrical change processing on the leading index and the consistent index respectively, so as to obtain the symmetric change rate C _w,i (t) of the leading index And consistent index symmetric rate of change C _{z, i} (t), the fifth sub-calculation module includes: the fourth sub-calculation module, used to carry out symmetric change processing to the leading index through the following formula, to obtain the leading index symmetric rate of change C _{w, i} (t):

in, is the number of the i (i=1, 2, ..., k _w ) leading indicators, t=2, 3, ..., n, k _w leading indicators; and the fifth sub-calculation module, used for the following formula The consistent index is processed symmetrically to obtain the symmetrical change rate C _z,i (t) of the consistent index:

其中，是第i(i＝1，2，…，k_z)个一致指标，t=2,3,…，n，k_z是一致指标的个数；第六子处理模块，用于对先行指标对称变化率C_w,i(t)和一致指标对称变化率C_z,i(t)进行标准化处理和趋势调整之后得到的结果进行合成计算，以获取先行合成指数和一致合成指数。

in, is the i-th (i=1, 2, ..., k _z ) consistent index, t=2, 3, ..., n, k _z is the number of consistent indicators; the sixth sub-processing module is used to be symmetrical to the leading index The results obtained after standardization and trend adjustment of the rate of change C _w,i (t) and the symmetrical rate of change of the consistent index C _z,i (t) are combined and calculated to obtain the leading composite index and the consistent composite index.

求对称变化率C_w,i(t)：Specifically, the fifth sub-processing module uses the exponential synthesis model to find the symmetrical rate of change of the index and standardizes it through the sixth sub-processing module, and the fourth sub-calculation module in the fifth sub-processing module according to the following formula

Find the symmetrical rate of change C _w,i (t):

t=2,3,…，n，其中，

是第i(i＝1，2，…，k_w)个先行指标，k_w是先行指标的个数。

t=2,3,...,n, where,

is the i-th (i=1, 2, . . . , k _w ) leading indicator, and k _w is the number of leading indicators.

Through the fifth sub-calculation module according to the following formula Find the symmetrical rate of change C _z,i (t):

t=2,3,…，n，其中，

是第i(i＝1，2，…，k_z)个一致指标，k_z是一致指标的个数。

t=2,3,...,n, where,

is the i-th (i=1, 2, ..., k _z ) consistent index, and k _z is the number of consistent indexes.

t=2,3,…,n；第七子处理模块，用于采用标准化因子A_w，i和A_z，i分别将先行指标对称变化率C_w,i(t)和一致指标对称变化率C_z,i(t)进行标准化处理，以得到标准化变化率S_w,i(t)和S_z,i(t)，其中，In the above-mentioned embodiment of the present application, the sixth sub-processing module may include: a sixth sub-calculation module, which is used to obtain the normalization factors A _{w, i} and A _{z, i} through the following formula:

t=2,3,...,n; the seventh sub-processing module is used to use normalization factors A _{w, i} and A _{z, i} to respectively convert the leading index symmetric rate of change C _w,i (t) and the consistent index symmetric rate of change C _z,i (t) is standardized to obtain standardized rate of change S _w,i (t) and S _z,i (t), where,

$S_{w,i}\left(t\right)=\frac{C_{w,i}\left(t\right)}{A_{w,i}},$ $S_{z,i}\left(t\right)=\frac{C_{z,i}\left(t\right)}{A_{z,i}},$ t=2,3,...,n;

The eighth sub-processing module is used to process the average rate of change on the standardized rate of change S _w,i (t) and S _z,i (t), so as to obtain the standardized average rate of change V _w (t) of the leading index and the consistent index The standardized average rate of change V _z (t); the seventh sub-calculation module is used to perform synthetic calculations based on the standardized average rate of change V _w (t) of the leading index and the standardized average rate of change V _z (t) of the consistent index, to Obtain the leading synthetic index I _w (t) and the consistent synthetic index I _z (t), where, $I_z\left(t\right)=I_z(t-1)\&times;\frac{200+V_z\left(t\right)}{200-V_z\left(t\right)},$ And I _w (1)=100, I _z (1)=100.

Specifically, according to the following formula, the sixth sub-calculation module performs standardized calculations on the symmetrical rate of change C _w,i (t) of the leading index and the symmetrical rate of change C _z,i (t) of the consistent index, so that the average absolute value is equal to 1, and the obtained Normalization factors A _w,i and A _z,i :

$A_{w,i}=\underset{t=2}{\overset{\mathrm{no}}{\mathrm{\&Sigma;}}}\frac{\left|C_{w,i}\left(t\right)\right|}{\mathrm{no}-1},$ $A_{z,i}=\underset{t=2}{\overset{\mathrm{no}}{\mathrm{\&Sigma;}}}\frac{\left|C_{z,i}\left(t\right)\right|}{\mathrm{no}-1},$ t=2,3,...,n,

Then the seventh sub-processing module standardizes C _{w, i} (t) and C z, i (t) according to A _{w, i} and A _{z , i} respectively, and obtains standardized rate of change S _{w, i} (t) and S _{z, i} (t):

t=2,3,…,n，之后第八子处理模块根据标准化变化率S_w,i(t)和S_z,i(t)进行平均变化率处理，以获取先行指标的标准化平均变化率V_w(t)和一致指标的标准化平均变化率V_z(t)，并通过第七子计算模块根据先行指标的标准化平均变化率V_w(t)和一致指标的标准化平均变化率V_z(t)进行合成计算，以获取先行合成指数I_w(t)和一致合成指数I_z(t)，具体地：令I_w(1)＝100，I_z(1)＝100，则

t=2,3,...,n, then the eighth sub-processing module performs average change rate processing according to standardized change rates S _w,i (t) and S _z,i (t) to obtain the standardized average change rate of leading indicators V _w (t) and the standardized average rate of change V _z (t) of the consistent index, and through the seventh sub-calculation module according to the standardized average rate of change V _w (t) of the leading index and the standardized average rate of change V _z ( t) Perform composite calculations to obtain the leading composite index I _w (t) and the consistent composite index I _z (t), specifically: let I _w (1)=100, I _z (1)=100, then

${\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&times;</span>\frac{\mathrm{<span\; class="notranslate">\; 200</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; V</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; 200</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; V</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; ,</span>$ ${\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&times;</span>\frac{\mathrm{<span\; class="notranslate">\; 200</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; V</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; 200</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; V</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; .</span>$

More specifically, after obtaining the leading composite index I _w (t) and the consistent composite index I _z (t), the power demand trend can be adjusted in the following way:

According to the following compound interest formula, calculate the respective average growth rate for each sequence of the consistent index group:

$r_i=(\sqrt[m_i]{C_{\mathrm{Li}}/C_{\mathrm{II}}}-1)\&times;100,$ i=1,2,…,k _z ,

和

分别是一致指标组第i个指标最先与最后循环的平均值，m_Ii与m_Li分别是一致指标组第i个指标最先与最后循环的月数，k₂是一致指标个数，m_i是最先循环的中心到最后循环的中心之间的月数。Wherein, Fig. 6 is a schematic diagram of trend adjustment according to the embodiment shown in Fig. 4, as shown in Fig. 6, t is the month,

and

are the average values of the first and last cycles of the i-th index of the consistent index group, m _Ii and m _Li are the months of the first and last cycles of the i-th index of the consistent index group, k ₂ is the number of consistent indicators, m _i is the number of months between the center of the first cycle to the center of the last cycle.

Then calculate the average growth rate G _r of the consistent index group and use it as the target trend: Then use the compound interest formula to calculate their respective average growth rates r′ _w and r′ _z for the initial composite indices I _w (t) and I _z (t) of the leading and consistent indicators:

${\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>\sqrt[{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}]{{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; Lw</span>}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; Iw</span>}}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&times;</span>\mathrm{<span\; class="notranslate">\; 100</span>}<span\; class="notranslate">\; ,</span>$ ${\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>\sqrt[{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}]{{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; Lz</span>}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; Iz</span>}}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&times;</span>\mathrm{<span\; class="notranslate">\; 100</span>}<span\; class="notranslate">\; ,</span>$

in,

Then make trend adjustments for the standardized average rate of change V _w (t) and V _z (t) of the leading index group and the consistent index group:

V′ _w (t)=V _w (t)+(G _r -r′ _w ), V′ _z (t)=V _z (t)+(G _r -r′ _z ), t=2,3, ..., n. Then calculate composite index according to the method in above-mentioned embodiment: make I' _w (1)=100, I' _z (1)=100, then

${{\mathrm{<span\; class="notranslate">\; I</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{{\mathrm{<span\; class="notranslate">\; I</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&times;</span>\frac{\mathrm{<span\; class="notranslate">\; 200</span>}<span\; class="notranslate">\; +</span>{{\mathrm{<span\; class="notranslate">\; V</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; 200</span>}<span\; class="notranslate">\; -</span>{{\mathrm{<span\; class="notranslate">\; V</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; ,</span>$ ${{\mathrm{<span\; class="notranslate">\; I</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{{\mathrm{<span\; class="notranslate">\; I</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&times;</span>\frac{\mathrm{<span\; class="notranslate">\; 200</span>}<span\; class="notranslate">\; +</span>{{\mathrm{<span\; class="notranslate">\; V</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; 200</span>}<span\; class="notranslate">\; -</span>{{\mathrm{<span\; class="notranslate">\; V</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; ,</span>$

Generate the leading composite index CI _w (t) and consistent composite index CI _z (t) with the base year as 100:

${\mathrm{<span\; class="notranslate">\; CI</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span>\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}<span\; class="notranslate">\; \&times;</span>\mathrm{<span\; class="notranslate">\; 100</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>$ ${\mathrm{<span\; class="notranslate">\; CI</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span>\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&times;</span>\mathrm{<span\; class="notranslate">\; 100</span>}<span\; class="notranslate">\; ,</span>$

和

分别是I′_w(t)和I′_z(t)在基准年份的平均值。in

and

are the mean values of I′ _w (t) and I′ _z (t) in the base year, respectively.

According to the above-mentioned embodiments of the present application, the eighth sub-processing module may include: a ninth sub-processing module, which is used to respectively convert the normalized rate of change S _w,i (t) of the leading index and the standardized rate of change S of the consistent index through the following formula _z,i (t) performs average rate of change processing to obtain the average rate of change R _w (t) of leading indicators and the average rate of change R _z (t) of consistent indicators:

$R_w\left(t\right)=\frac{\underset{i=1}{\overset{k_w}{\mathrm{\&Sigma;}}}{S}_{w,i}\left(t\right){\mathrm{\&lambda;}}_{w,i}}{\underset{i=1}{\overset{k_w}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_{w,i}},$ $R_z\left(t\right)=\frac{\underset{i=1}{\overset{k_z}{\mathrm{\&Sigma;}}}{S}_{z,i}\left(t\right){\mathrm{\&lambda;}}_{z,i}}{\underset{i=1}{\overset{k_z}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_{z,i}},$ Among them, λw _,i and λz _,i are the weights of the leading indicator and the i-th indicator of the consistent indicator respectively; the eighth sub-calculation module is used to obtain the indicator standardization factor F _w through the following formula:

${\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; [</span>\underset{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 2</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; |</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; [</span>\underset{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 2</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; |</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; ;</span>$

The ninth sub-calculation module is used to process the standardized average rate of change according to the index normalization factor _Fw , so as to obtain the standardized average rate of change _Vw (t) of the leading index and the standardized average rate of change _V2 (t) of the consistent index, where , V _w (t)=R _w (t)/F _w , V _z (t)=R _z (t).

Specifically, the standardized rate of change S _w,i (t) of the leading indicator and the standardized rate of change S _z,i (t) of the consistent indicator are processed by the average rate of change respectively by the following formula to obtain the average rate of change R of the leading indicator _w (t) and the average rate of change R _z (t) of the consensus indicator:

$R_w\left(t\right)=\frac{\underset{i=1}{\overset{k_w}{\mathrm{\&Sigma;}}}{S}_{w,i}\left(t\right){\mathrm{\&lambda;}}_{w,i}}{\underset{i=1}{\overset{k_w}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_{w,i}},$ $R_z\left(t\right)=\frac{\underset{i=1}{\overset{k_z}{\mathrm{\&Sigma;}}}{S}_{z,i}\left(t\right){\mathrm{\&lambda;}}_{z,i}}{\underset{i=1}{\overset{k_Z}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_{z,i}},$ λ _w,i and λ _z,i are the weights of the i-th index of the leading and consistent index groups, respectively.

Then calculate the exponential normalization factor F _w according to the following formula:

${\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; [</span>\underset{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 2</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; |</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; [</span>\underset{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 2</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; |</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; ;</span>$

Finally, the normalized average rate of change is processed according to the index standardization factor F _w to obtain the standardized average rate of change V _w (t) of the leading index and the standardized average rate of change V _z (t) of the consistent index:

V _w (t) = R _w (t)/F _w , V _z (t) = R _z (t), t = 2, 3,..., n, where the amplitude of the average rate of change of the consistent index sequence is used to The purpose of adjusting the average rate of change of the leading index series and the lagging index series is to apply the two indexes as a coordinated system.

According to the above-mentioned embodiments of the present application, after executing the acquisition module 10, the device may further include: a tenth sub-processing module, for preprocessing the data in the data sequence, the preprocessing includes: filling missing data processing, correcting noise data processing, data smoothing, and data normalization.

Fig. 3 is a flow chart of a method for obtaining early warning parameters of power demand according to an embodiment of the present invention. Fig. 4 is a detailed flow chart of a method for obtaining early warning parameters of power demand according to an embodiment of the present invention.

As shown in Figure 3 and Figure 4, the method includes the following steps:

Step S102, acquiring a data sequence for generating early warning indicators.

Wherein, this step can be realized by step S202 in FIG. 4 : collecting monthly data on macroeconomics and power demand, and performing preprocessing.

In step S104, the data series are screened according to the adjustment parameters to obtain a data series including trend items and period items. Wherein, this method can be implemented through step S204 in FIG. 4: by making seasonal adjustments to the data series, the trend items and period items of all indicators are obtained.

Specifically, in the field of electricity demand, the data sequence for generating early warning indicators includes four parts: long-term trend items (ie trend items), cycle items caused by periodic laws such as industry expansion-prosperity contraction-recession-re-expansion (ie cycle items) Seasonal items caused by different seasons, spring, summer, autumn and winter or the number of days per month, random items caused by unknown factors, seasonal adjustments are made to the indicators of data value and physical quantities, seasonal items and random items in the data sequence are eliminated, and trend items are obtained and a data sequence of periodic terms.

Step S106, calculate the trend index of the data sequence containing trend items and period items, and filter the data series containing trend items and period items according to the trend index to obtain the early warning index sequence, which is the trend in the data sequence Exponential is the growth data. Among them, this step can be realized through steps S206 to S208: step S206 judges whether the index data in the data sequence is growth index index data, and if yes, executes step S208: calculate the growth rate of the index, and in the data sequence In the case that the index data is not growth index index data, step S210 is executed.

Step S108, extracting the power generation amount in the early warning index sequence as the reference index, and extracting the indexes other than the power generation amount as the selected indexes. Wherein, the method is realized through step S210: determining a benchmark index.

Specifically in the field of electricity demand, quarterly GDP or monthly industrial added value is used as the benchmark indicator.

Step S110, perform correlation calculation on the warning indicators according to the time difference analysis model and/or the K-L information model, so as to obtain the correlation coefficient between each selected index and the benchmark index, and filter the selected indicators according to the correlation coefficient , to get the leading and consistent indicators. Among them, step S212 can implement this method: use the time difference analysis model to screen the initial consistent indicators and initial leading indicators; then perform step S214: use the K-L information model to screen the final consistent indicators and leading indicators from the primary consistent indicators and leading indicators .

Specifically, the time-difference analysis model can be used to preliminarily screen consistent indicators and leading indicators from indicators such as macroeconomics and industrial product output.

In step S112, the leading index and the consistent index are synthesized according to the synthetic index model, so as to obtain the leading synthetic index and the consistent synthetic index as early warning parameters. Among them, the above method is realized by executing step S216: using the index synthesis model to synthesize the consistent index and the leading index into the consistent index and the leading index; Demand fluctuates periodically.

Specifically, use the index synthesis model to synthesize several leading indicators into a leading index, and synthesize several consistent indicators into a consistent index. After obtaining the early warning parameters, the future of the consistent indicators can be judged based on the early warning parameters. purpose of the trend.

Using the method for obtaining early warning parameters of electric power demand of the present invention, the data sequence obtained for generating early warning indicators is screened, and the data sequence containing trend items and period items is obtained, and then the trend index of the above data sequence is calculated, and Filter the data sequence according to the trend index to obtain the early warning index sequence in which the trend index is increasing in the data sequence, then extract the power generation in the early warning index sequence as the benchmark index, and extract the indicators other than electricity consumption as the selected indicators , and then calculate the correlation of the early warning indicators according to the time difference analysis model and/or the K-L information model, so as to obtain the correlation coefficient between each selected index and the benchmark index, and filter the selected indicators according to the correlation coefficient, In order to obtain the leading index and the consistent index, finally, according to the composite index model, the leading index and the consistent index are synthesized to obtain the leading composite index and the consistent composite index as early warning parameters. Through the method of obtaining early warning parameters of electric power demand in this application, the trend items and period items in the data sequence are obtained first, and then the leading indicators and consistent indicators are obtained through screening, and the above leading indicators and consistent indicators are synthesized with an index synthesis model to obtain early warning Parameters, which solves the problem in the prior art that using forecasting technology to obtain short-term periodic fluctuation distortion in the field of power demand, and the early warning parameters that meet the power demand cannot be obtained, resulting in the inability to formulate reasonable countermeasures according to the periodic fluctuation of power. Realized Accurately obtain the early warning parameters of power demand, so as to accurately formulate reasonable measures to deal with periodic fluctuations according to the effect of short-term periodic fluctuations, thereby slowing down the magnitude of periodic fluctuations and reducing the damage caused by periodic fluctuations to the power industry and economic development.

In the above-mentioned embodiment of the present invention, the data sequence is firstly screened through step S102, that is, the original data sequence is seasonally adjusted, that is, the influence of seasonal factors and random factors is removed, and the long-term trend item (that is, in the above-mentioned embodiment Based on the trend item) and the short-term cycle item (that is, the cycle item in the above-mentioned embodiment), through index screening and index synthesis, the fluctuation trend of power demand is studied with the boom early warning technology, and the above-mentioned embodiment uses the economic early warning power demand, Select leading indicators and consistent indicators from a large number of economic indicators to judge the periodic fluctuations of power demand, making the judgment results more accurate and making the periodic fluctuations used to formulate reasonable countermeasures more accurate and reasonable.

Wherein, the preprocessing in step S202 may include: filling missing data processing, correcting noise data processing, data smoothing processing and data normalization processing, and the processed data sequence is taken as Y ⁽⁰⁾ .

Specifically, in the above-mentioned embodiments of the present application, before performing step S204, the method further includes the following steps:

(1) Remove the difference in the number of working days in a given month between different years caused by holidays or other reasons. The specific method is as follows:

(t＝1,2，…,n)共有m年，n个月，n=m×12，t代表月份数，i代表星期数，j代表年数，则各月实际工作天数为D_t(t=1，2，…,n)，m年平均的每个月份的工作天数为：

(L=1,2,…,12)，可得到工作天数调整系数序列p_t：In this example, suppose the monthly data series to be analyzed

(t=1,2,...,n) there are m years, n months, n=m×12, t represents the number of months, i represents the number of weeks, j represents the number of years, then the actual number of working days in each month is D _t (t =1, 2,...,n), the average number of working days in each month in m years is:

(L=1,2,…,12), the working days adjustment coefficient sequence p _t can be obtained:

t=1，2，…,n，然后根据工作天数调整系数p_t得到月份调整后的月度数据序列Y⁽¹⁾： $Y_t^{\left(1\right)}=Y_t^{\left(0\right)}/p_t.$

t=1, 2,...,n, and then adjust the coefficient p _t according to the number of working days to get the adjusted monthly data series Y ⁽¹⁾ : $Y_t^{\left(1\right)}=Y_t^{\left(0\right)}/p_t.$

(2) Adjust the working day in the sequence: extract the changes caused by the different working days (week structure) of each month from the original sequence.

Assume that the variable elements of the working day of the week are included in the irregular elements, that is, the form of the irregular elements is ID _r , assuming that ID _r has been decomposed from the original sequence, use regression analysis to find the corresponding weights of Monday, Tuesday, ..., days , so that the ID _r is decomposed into the real irregular element I and the weekly working day variable element D _r .

${\mathrm{<span\; class="notranslate">\; ID</span>}}_{\mathrm{<span\; class="notranslate">\; rt</span>}}\text{<span class="notranslate"> -1.0=</span>}\frac{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}\mathrm{<span\; class="notranslate">\; <figure-callout\; id="1"\; label="t\; B"\; filenames="HDA00002333373000011.png"\; state="\{\{state\}\}">t</figure-callout></span><figure-callout\; id="1"\; label="t\; B"\; filenames="HDA00002333373000011.png"\; state="\{\{state\}\}">\; </figure-callout>}}{\mathrm{<span\; class="notranslate">\; </span>}}_{}{\mathrm{<span\; class="notranslate">B</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; <figure-callout\; id="2"\; label="t\; B"\; filenames="HDA00002333373000011.png"\; state="\{\{state\}\}">t</figure-callout></span><figure-callout\; id="2"\; label="t\; B"\; filenames="HDA00002333373000011.png"\; state="\{\{state\}\}">\; </figure-callout>}}{\mathrm{<span\; class="notranslate">\; </span>}}_{}{\mathrm{<span\; class="notranslate">B</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; \&CenterDot;</span><span\; class="notranslate">\; \&CenterDot;</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; 7</span>}\mathrm{<span\; class="notranslate">\; t</span>}}{\mathrm{<span\; class="notranslate">\; B</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}}{{\mathrm{<span\; class="notranslate">\; A</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; ,</span>$

)；A_t为t月的天数，2月取28.25天；I_t为真正的不规则要素。In the above formula: ID _rt is an irregular element including the variable element D _r of weekday in month t; x _it is the number of days in week i in month t (t=1,...,n); B _i is the number of days in week i Weights(

); A _t is the number of days in month t, 28.25 days in February; I _t is the real irregular element.

When the estimated value of _Bi thus obtained is _bi , the variable element D _r of the weekly working day in month t can be calculated according to the following formula:

D _rt ={x _1t (b ₁ +1)+x _2t (b ₂ +1)+...+x _7t (b ₇ +1)}/A _t .

(3) After adjusting the data series according to factors of holidays and working days, correct the peculiar items in the series.

When decomposing various factors in the economic time series, it is necessary to pre-correct items with significant outliers in irregular changes (that is, idiosyncratic items, such as strikes, the impact of bad weather, data errors, etc.). The method is:

a. The setting of the threshold value of the unique item.

即：Assume that the irregular element I has been decomposed from the original sequence. In order to exclude outliers in the irregular variable element I, it is necessary to calculate the standard deviation of the 5-year moving average of I. First calculate the initial 5-year moving average standard deviation

Right now:

${\mathrm{\&sigma;}}_{\mathrm{<figure-callout\; id="0"\; label="j"\; filenames="HDA00002333373000011.png"\; state="\{\{state\}\}">j</figure-callout>}}^0=\sqrt{\frac{1}{60}\underset{t=j\&times;12-36+1}{\overset{j\&times;12+\mathrm{twenty\; four}}{\mathrm{\&Sigma;}}}{(I_t-{\stackrel{\&OverBar;}{I}}_j)}^2},$ j=3,4,...,m-2

对应于5年期间的中心年，每年计算出一个

故

是一个年度序列。可以认为满足

的I_t是特异的，除去这些I_t，由下式：In the above formula is the 5-year moving average of the I series, m is the number of years in the I series, t=1,2,...,n (t is the number of months in the I series). let

Corresponding to the central year of the 5-year period, one is calculated for each year

so

is an annual sequence. can be considered satisfied

The I _t is specific, remove these I _t , by the following formula:

${\mathrm{\&sigma;}}_j=\sqrt{\frac{1}{60-a}\mathrm{\&Sigma;}{(I_t-{\stackrel{\&OverBar;}{I}}_j)}^2},$ j=3,4,...,m-2,

Calculate the 5-year moving standard deviation {σ _j } again, where t=1,2,…,n (t is the number of months in the I sequence), and a is the number of outliers. Two items are missing at both ends of the {σ _j } sequence, and the {σ _j } of the 3rd year from the start and end are used to replace the missing two-year σ _j values at both ends.

b. The correction of the peculiar items.

The modified weight w is calculated according to the 5-year moving standard deviation {σ _j },

替换I_t。修正特异项后的I序列记为I^w。 Among them, t=1,2,…,n (t is the number of months in the I sequence), j=1,2,…,m, using the above inequalities can correct the peculiar term of the I sequence: It corresponds to w _t <1 , with this w _t as the weight, the two items of I _t-2 , I _t-1 , I _t+1 , I _t+2 that are similar to it before and after (note that the w corresponding to the selected item must be equal to 1 , otherwise take the value next to it) as a weighted average of 5 items, and use the value obtained in this way Replace I _t . If I _t corresponding to w _t <1 is located at both ends, take this w _t as the weight, and a total of 4 items of the I value of the three items w _t = 1 close to it are used as weighted average, and the obtained average value is used

Replace I _t . The I sequence after correcting the specific items is recorded as I ^w .

进行初始估计，获取初始趋势循环成分。(4) According to the monthly data series

Make an initial estimate to obtain the initial trend cycle component.

Estimating the Cyclic Component of Trend in a Series Using a Centered 12-Term Moving Average

${\mathrm{<span\; class="notranslate">\; TC</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; (</span>\frac{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 6</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 5</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; \&CenterDot;</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; \&CenterDot;</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; \&CenterDot;</span><span\; class="notranslate">\; \&CenterDot;</span><span\; class="notranslate">\; \&CenterDot;</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 5</span>}}}{\mathrm{<span\; class="notranslate">\; 12</span>}}<span\; class="notranslate">\; +</span>\frac{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 5</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; \&CenterDot;</span><span\; class="notranslate">\; \&CenterDot;</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; \&CenterDot;</span><span\; class="notranslate">\; \&CenterDot;</span><span\; class="notranslate">\; \&CenterDot;</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 5</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 6</span>}}}{\mathrm{<span\; class="notranslate">\; 12</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>$

$<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; twenty\; four</span>}}{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 6</span>}}<span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 12</span>}}{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 5</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; \&CenterDot;</span><span\; class="notranslate">\; \&CenterDot;</span><span\; class="notranslate">\; \&CenterDot;</span><span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 12</span>}}{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; \&CenterDot;</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 12</span>}}{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 5</span>}}<span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; twenty\; four</span>}}{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 6</span>}}$

中的元素。Among them, y _t-6 , y _t-5 , …, y _t , …, y _t+5 , y _t+6 are monthly data series

elements in .

估计季节不规则成分

(5) Cycle ingredients according to trend

Estimated Seasonal Irregularity Component

对每个月份应用3×3移动平均初步估计季节成分：(6) Irregular composition according to season

Apply a 3×3 moving average to each month to initially estimate the seasonal component:

First, according to the following formula, the seasonal factor is obtained

${\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; S</span>}}}_{\mathrm{<span\; class="notranslate">\; t</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 3</span>}}<span\; class="notranslate">\; (</span>\frac{{\mathrm{<span\; class="notranslate">\; Si</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; twenty\; four</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; Si</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 12</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; Si</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}}{\mathrm{<span\; class="notranslate">\; 3</span>}}<span\; class="notranslate">\; +</span>\frac{{\mathrm{<span\; class="notranslate">\; Si</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 12</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; Si</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; Si</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 12</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}}{\mathrm{<span\; class="notranslate">\; 3</span>}}<span\; class="notranslate">\; +</span>\frac{{\mathrm{<span\; class="notranslate">\; Si</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; Si</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 12</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; Si</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; twenty\; four</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}}{\mathrm{<span\; class="notranslate">\; 3</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ;</span>$

$\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 9</span>}}{\mathrm{<span\; class="notranslate">\; Si</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; twenty\; four</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; 9</span>}}{\mathrm{<span\; class="notranslate">\; Si</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 12</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 3</span>}}{\mathrm{<span\; class="notranslate">\; 9</span>}}{\mathrm{<span\; class="notranslate">\; Si</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; 9</span>}}{\mathrm{<span\; class="notranslate">\; Si</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 12</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 9</span>}}{\mathrm{<span\; class="notranslate">\; Si</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; twenty\; four</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}$

以使得因子之和在每一个连续的12个月内都近似为零：Then the seasonal factor is standardized and calculated to obtain the standard seasonal factor

so that the factors sum to approximately zero for each consecutive 12-month period:

${\mathrm{<span\; class="notranslate">\; S</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; =</span>{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; S</span>}}}_{\mathrm{<span\; class="notranslate">\; t</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; twenty\; four</span>}}{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; S</span>}}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 6</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 12</span>}}{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; S</span>}}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 5</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 12</span>}}{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; S</span>}}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 5</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; twenty\; four</span>}}{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; S</span>}}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 6</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; .</span>$

初次估计季节调整后序列

(7) According to the standard seasonal factor

Initially estimated seasonally adjusted series

(8) Use the 13-item moving average to further estimate the trend cycle component

${\mathrm{<span\; class="notranslate">\; TC</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 16796</span>}}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; 375</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 6</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; 468</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 5</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 1100</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 3</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 2475</span>}\mathrm{<span\; class="notranslate">\; TC</span>}{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 2</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 3600</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}$

$<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 4032</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 3600</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 2475</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 2</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 1100</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 3</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; 468</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 5</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; 325</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 6</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; .</span>$

(9) Further estimate seasonal irregularities:

(10) Estimate the final seasonal component with a 3×5 moving average:

Obtain the final seasonal factor according to the following formula

${\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; S</span>}}}_{\mathrm{<span\; class="notranslate">\; t</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 15</span>}}{\mathrm{<span\; class="notranslate">\; Si</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 36</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; 15</span>}}{\mathrm{<span\; class="notranslate">\; Si</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; twenty\; four</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 3</span>}}{\mathrm{<span\; class="notranslate">\; 15</span>}}{\mathrm{<span\; class="notranslate">\; Si</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 12</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 3</span>}}{\mathrm{<span\; class="notranslate">\; 15</span>}}{\mathrm{<span\; class="notranslate">\; Si</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 3</span>}}{\mathrm{<span\; class="notranslate">\; 15</span>}}{\mathrm{<span\; class="notranslate">\; Si</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 12</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; 15</span>}}{\mathrm{<span\; class="notranslate">\; Si</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; twenty\; four</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 15</span>}}{\mathrm{<span\; class="notranslate">\; Si</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 36</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}$

进行标准化得到标准化的最终季节因子使得因子之和在每一个连续的12个月内都近似为零：Then for the seasonal factor

Normalize to get the final standardized seasonal factor Make the factors sum to approximately zero for each consecutive 12-month period:

${\mathrm{<span\; class="notranslate">\; S</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; =</span>{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; S</span>}}}_{\mathrm{<span\; class="notranslate">\; t</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; twenty\; four</span>}}{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; S</span>}}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 6</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 12</span>}}{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; S</span>}}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 5</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 12</span>}}{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; S</span>}}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 5</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; twenty\; four</span>}}{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; S</span>}}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 6</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; .</span>$

估计季节调整后的序列

(11) According to the standardized final seasonal factor

Estimating the Seasonally Adjusted Series

(12) Obtain the final trend cycle component, the method is as follows:

First estimate the irregular component:

和趋势项的月增长率的绝对值之和的比值来衡量不规则成分的显著性

Then use the irregular term

and trending items The ratio of the sum of the absolute values of the monthly growth rates to measure the significance of the irregular components

$\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; I</span>}}<span\; class="notranslate">\; /</span>\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; C</span>}}<span\; class="notranslate">\; =</span>\frac{\underset{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 2</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; /</span>{{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; |</span>}{\underset{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 2</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; TC</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}{{\mathrm{<span\; class="notranslate">\; TC</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; |</span>}<span\; class="notranslate">\; ,</span>$

则用下述9项移动平均估计最终的趋势循环成分：if

Then use the following 9-term moving average to estimate the final trend cycle component:

${\mathrm{<span\; class="notranslate">\; TC</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 2431</span>}}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; 99</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 4</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; twenty\; four</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 3</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 288</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 2</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 348</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 805</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; ;</span>$

$<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 648</span>}{\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 288</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 2</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; twenty\; four</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 3</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; 99</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 4</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; )</span>$

则用下述13项移动平均估计最终的趋势循环成分：if

The final trend cycle component is then estimated using the following 13-term moving average:

${\mathrm{<span\; class="notranslate">\; TC</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 16796</span>}}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; 325</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 6</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 468</span>}\mathrm{<span\; class="notranslate">\; TC</span>}{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 5</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 1100</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 3</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 2475</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 2</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 3600</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; ;</span>$

$<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 4032</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 3600</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 2475</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 2</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 1100</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 3</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; 468</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 5</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; 325</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 6</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; )</span>$

则用下述23项移动平均估计最终的趋势循环成分：if

Then use the following 23-term moving average to estimate the final trend cycle component:

${\mathrm{<span\; class="notranslate">\; TC</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 4032015</span>}}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; 17250</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 11</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; 44022</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 10</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; 63250</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 9</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; 5757</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 8</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; 19950</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 7</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; )</span>$

$<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 54150</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 6</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 156978</span>}{\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 5</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 275400</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 4</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 392700</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 3</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 491700</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 2</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 557700</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; ,</span>$

$<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 580853</span>}{\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 557700</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 491700</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 2</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 392700</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 3</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 275400</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 4</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 156978</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 5</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}$

$<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 54150</span>}{\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 6</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; 19950</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 7</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; 58575</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 8</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; 63250</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 9</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; 44022</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 10</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; 17250</span>}\mathrm{<span\; class="notranslate">\; TCI</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 11</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; )</span>$

The final trend cycle component is obtained through the above calculation

和季节调整后的序列

估计最终的不规则成分： $I_t^{\left(3\right)}={\mathrm{TCI}}_t^{\left(2\right)}-{\mathrm{TC}}_t^{\left(3\right)}.$ (13) Cycle ingredients according to trend

and the seasonally adjusted series

Estimate the final irregular composition: $I_t^{\left(3\right)}={\mathrm{TCI}}_t^{\left(2\right)}-{\mathrm{TC}}_t^{\left(3\right)}.$

Then the sequence Y ⁽²⁾ =TC ⁽³⁾ containing only trend cycle items after seasonal adjustment is obtained.

是发展指数的指标，不做变换，即Y⁽³⁾=Y⁽²⁾。In the above-mentioned embodiment of the present application, step S108 in FIG. 4 is specifically realized by the following method: for the physical quantity and the value quantity index, calculate the development index of the sequence after seasonal adjustment:

is the indicator of the development index, without transformation, that is, Y ⁽³⁾ =Y ⁽²⁾ .

In the above-mentioned embodiments of the present application, step S110 is performed: performing correlation calculation on the early warning indicators according to the time difference analysis model and/or the K-L information model, so as to obtain the correlation coefficient between each selected indicator and the benchmark indicator, and The process of screening the selected indicators according to the correlation coefficient to obtain leading indicators and consistent indicators includes the following steps:

Obtain the correlation coefficient r _l between each selected index and the benchmark index according to the following formula:

Among them, l=0,±1,±2,…,±12, Y=(y ₁ ,y ₂ ,…,y _n ) is the benchmark index, X=(x ₁ ,x ₂ ,…,x _n ) is Selected indicators, l is the time difference, specifically the leading or lagging period, n _l is the number of all indicators; the selected indicators whose time difference is within the first value range and the correlation coefficient r _l is greater than the first threshold As the leading index, the selected index whose time difference value is within the second value range and whose correlation coefficient r _l is greater than the second threshold is taken as the consistent index. Wherein, the first value range may be less than -3, the first threshold may be 0.7, the second value range may be greater than or equal to -2 and less than or equal to 2, and the second threshold may be 0.7.

Specifically, in this step, benchmark indicators are used as the "benchmark" for screening, and the time difference correlation analysis model is used to initially screen consistent indicators and leading indicators. All the selected indicators except the benchmark indicators are ahead or lagged by l period (i.e. the time difference in the above-mentioned embodiment) (l=0, ±1, ±2,..., ±12), and each selected index is calculated according to the following formula The correlation coefficient r _l between the index and the benchmark index:

$r_l=\frac{\underset{t=1}{\overset{{\mathrm{no}}_l}{\mathrm{\&Sigma;}}}(x_{t+l}-\stackrel{\&OverBar;}{x})({\mathrm{the\; y}}_t-\stackrel{\&OverBar;}{\mathrm{the\; y}})}{\sqrt{\underset{t=1}{\overset{{\mathrm{no}}_l}{\mathrm{\&Sigma;}}}{(x_{t+l}-\stackrel{\&OverBar;}{x})}^2{({\mathrm{the\; y}}_t-\stackrel{\&OverBar;}{\mathrm{the\; y}})}^2}},$ l=0, ±1, ±2, ..., ±12,

和

分别为序列X和Y的平均值。n_l是所有指标的数据个数。则最大的时差相关系数被认为反映了被选指标与基准指标的时差相关关系，相应的延迟数l表示超前或滞后期，即使得时差相关性系数r_l最大的延迟数l就是该被选指标与基准指标的超前或滞后期。In the above formula, Y=(y ₁ ,y ₂ ,…,y _n ) is the benchmark index, X=(x ₁ ,x ₂ ,…,x _n ) is the selected index, r is the correlation coefficient, l means leading or lagging Period (instantaneous difference), when l is negative, it means ahead, when it is positive, it means lag, l is called time difference or delay number,

and

are the mean values of sequences X and Y, respectively. n _l is the number of data of all indicators. Then the largest time-difference correlation coefficient is considered to reflect the time-difference correlation relationship between the selected index and the benchmark index, and the corresponding delay number l represents the leading or lagging period, that is, the largest delay number _l of the time-difference correlation coefficient r l is the selected index The lead or lag period with respect to the benchmark metric.

Fig. 5 is a schematic diagram of a method for obtaining a consistent index and a leading index according to the embodiment shown in Fig. 4. Specifically, after obtaining the correlation coefficient r _l , the leading index and the consistent index are obtained by the method shown in Fig. 5:

Step S302: Detect whether the index data is time difference<-3 and correlation coefficient>0.7 or whether the index data is -2≤time difference≤2 and correlation coefficient>0.7. Among them, when the time difference of the index data<-3 and the correlation coefficient>0.7 or the index data meets the condition -2≤time difference≤2 and the correlation coefficient>0.7, execute step S304, and if the index data does not meet the time difference<-3 And if the correlation coefficient is >0.7, and if -2≤time difference≤2 and the correlation coefficient>0.7 are not met, execute step S310: discard the index data.

Step S304: Obtain the initial leading index and the initial consistent index, wherein, when the time difference of the index data is <-3 and the correlation coefficient is >0.7, the data is selected as the initial leading index, and when the index data is -2≤time difference≤2 And when the correlation coefficient is >0.7, the index data is selected as the initial consistent index.

In the above-mentioned embodiment of the present application, according to the time difference analysis model and/or the KL information amount model, the correlation calculation is performed on the early warning indicators, and the correlation coefficient between each selected index and the benchmark index is used, and the correlation coefficient is calculated according to the correlation coefficient. The step of screening the selected indicators to obtain the leading indicators and consistent indicators includes: obtaining the correlation coefficient r _l between each selected indicator and the benchmark indicator according to the following formula:

其中，l=0，±1，±2，…，±12，Y＝(y₁,y₂,…,y_n)为基准指标，X=(x₁,x₂,…,x_n)为被选择指标，l为时差，n_l为所有指标的个数；将时差取值在第一取值范围内且相关性系数r_l大于第一阈值的被选择指标作为先行指标的初始指标，并将时差取值在第二取值范围内且相关性系数r_l大于第二阈值的被选择指标作为一致指标的初始指标；对基准指标、先行指标的初始指标以及一致指标的初始指标进行标准化处理，以获取标准基准指标序列p_t、标准被选择指标的序列q_t，其中，标准被选择指标包括标准先行指标以及标准一致指标；按如下公式获取每个标准被选择指标和标准基准指标之间的K-L信息量k_l：

Among them, l=0,±1,±2,…,±12, Y=(y ₁ ,y ₂ ,…,y _n ) is the benchmark index, X=(x ₁ ,x ₂ ,…,x _n ) is Selected indicators, l is the time difference, n _l is the number of all indicators; the selected index whose time difference value is within the first value range and the correlation coefficient r _l is greater than the first threshold is used as the initial indicator of the leading indicator, and The selected index whose time difference value is within the second value range and the correlation coefficient r _l is greater than the second threshold is used as the initial index of the consistent index; standardize the initial index of the benchmark index, the leading index, and the initial index of the consistent index , to obtain the standard benchmark index sequence p _t and the standard selected index sequence q _t , where the standard selected index includes the standard leading index and the standard consistent index; according to the following formula to obtain the relationship between each standard selected index and the standard benchmark index The amount of KL information k _l :

t=1，2，…，n，l为时差，n_l为所有指标的个数；将时差取值在第三取值范围内且K-L信息量k_l小于第三阈值的标准被选择指标作为先行指标，并将时差取值在第四取值范围内且K-L信息量k_l小于第四阈值的被选择指标作为一致指标。其中，第三取值范围可以是小于-3，第三阈值可以是0.3，第四取值范围可以是大于等于-2且小于等于2，第四阈值可以是0.3。k _l =∑p _t ln(p _t /q _t+1 ), where l=0，±1，…,±12，

t=1, 2,..., n, l is the time difference, n _l is the number of all indicators; the standard selected index that the time difference value is within the third value range and the KL information volume k _l is less than the third threshold is taken as The leading index, and the selected index whose time difference value is within the fourth value range and whose KL information volume k _l is less than the fourth threshold is taken as a consistent index. Wherein, the third value range may be less than -3, the third threshold may be 0.3, the fourth value range may be greater than or equal to -2 and less than or equal to 2, and the fourth threshold may be 0.3.

The above method can be realized through the following steps: standardize the benchmark index, and record the processed standard benchmark index sequence as p _t :

$p_t={\mathrm{the\; y}}_t/\underset{t=1}{\overset{\mathrm{no}}{\mathrm{\&Sigma;}}}{\mathrm{the\; y}}_t,$ t=1,2,...,n,

The leading indicators and consistent indicators of primary selection are also standardized, and the processed standard is recorded as q _t by the sequence of selected indicators:

$q_t=x_t/\underset{t=1}{\overset{\mathrm{no}}{\mathrm{\&Sigma;}}}{x}_t,$ t=1,2,...,n,

Then, the amount of KL information k _l about the benchmark index after the delay l of each primary index is calculated according to the following formula:

k _l =∑p _t ln(p _t /q _t+l )＝0，±1，…,±12

Among them, l represents the leading or lagging period. When l is a negative number, it means leading, and when it is positive, it means lagging. _l is called the time difference. The t in the formula represents the month, and the i represents the week number.

其相对应的延迟数l^*就是被选指标最适当的超前或滞后月数（季度），其中，K-L信息量越接近于0，说明指标x与基准指标y越接近。After calculating 2L+1 KL information volume, select a minimum value k _l′ from the k _l value as the KL information volume of the selected index x with respect to the benchmark index y, that is

The corresponding delay number l ^* is the most appropriate leading or lagging months (quarters) of the selected index, among which, the closer the KL information is to 0, the closer the index x is to the benchmark index y.

Specifically, the leading indicators and consistent indicators can be screened through the following steps, which are recorded as W ⁽³⁾ and Z ⁽³⁾ respectively:

Step S306: Detect whether the initial leading index is time difference<-3 and K-L information content<0.3 or whether the initial consistent data is -2≤time difference≤2 and K-L information content<0.3. Among them, when the initial leading index meets the time difference<-3 and the K-L information content<0.3 or the initial consistent data meets the condition that -2≤time difference≤2 and the K-L information content<0.3, step S308 is executed. -3 and K-L information volume<0.3, perform step S310: discard index data, and perform step S310: discard index data if the initial consistent data does not meet -2≤time difference≤2 and K-L information volume<0.3.

Step S308: Select a leading index and a consistent index. Among them, in the case of the time difference of the initial leading index <-3 and the K-L information volume <0.3, the index is selected as the current index. In the case of the initial consistent index data, -2≤time difference≤2 and the K-L information volume<0.3 , select this indicator as the consistent indicator. For example: select the index with delay number l=-2, -3, -4 as the leading index for primary selection, and the index with delay number l=-1, 0, 1 as the consistent index for primary selection.

According to the above-mentioned embodiment of the present application, the step of synthesizing the leading index and the consistent index according to the composite index model to obtain the leading composite index and the consistent composite index as early warning parameters includes: performing symmetrical change processing on the leading index and the consistent index respectively, In order to obtain the symmetrical rate of change C _{w, i} (t) of the leading index and the symmetrical rate of change C _{z, i} (t) of the consistent index, where the symmetrical change of the leading index is processed by the following formula to obtain the symmetrical rate of change C _w of the leading index _{, i} (t):

其中，

是第i(i＝1，2，…，k_w)个先行指标，t=2,3,…，n，k_w先行指标的个数；通过如下公式对一致指标进行对称变化处理，以获取一致指标对称变化率C_z,i(t)：

in,

is the number of the i-th (i=1, 2, ..., k _w ) leading indicators, t = 2, 3, ..., n, k _w leading indicators; through the following formula, carry out symmetrical change processing on the consistent indicators to obtain Symmetric rate of change of consistency index C _z,i (t):

其中，

是第i(i=1，2，…，k_z)个一致指标，t=2,3,…,n，k_z是一致指标的个数；对先行指标对称变化率C_w,i(t)和一致指标对称变化率C_z,i(t)进行标准化处理和趋势调整之后得到的结果进行合成计算，以获取先行合成指数和一致合成指数。其中，上述公式中的t代表月份。

in,

is the i-th (i=1, 2,...,k _z ) consistent index, t=2,3,...,n, k _z is the number of consistent indexes; the symmetric change rate C _w,i (t ) and the consistent index symmetric rate of change C _z,i (t) are standardized and trend-adjusted, and the results obtained after the synthetic calculation are performed to obtain the leading synthetic index and the consistent synthetic index. Wherein, t in the above formula represents a month.

和

求对称变化率C_w,i(t)和C_z,i(t)：Specifically, use the exponential synthesis model to find the symmetrical rate of change of the index and standardize it, and use the following formulas to calculate

and

Find the symmetrical rates of change C _w,i (t) and C _z,i (t):

$C_{w,i}\left(t\right)=200\&times;\frac{W_i^{\left(3\right)}\left(t\right)-W_i^{\left(3\right)}(t-1)}{W_i^{\left(3\right)}\left(t\right)+W_i^{\left(3\right)}(t-1)},$ t=2,3,...,n,

t=2,3,…,n，其中，

是第i(i＝1，2，…，k_w)个先行指标，

是第i(i＝1，2，…，k_z)个一致指标，k_w和k_z分别是先行指标和一致指标的个数。

t=2,3,…,n, where,

is the i (i=1, 2, ..., k _w ) leading index,

is the i-th (i=1, 2, ..., k _z ) consistent index, and k _w and k _z are the numbers of the leading index and the consistent index respectively.

t=2,3,…,n；采用标准化因子A_w，i和A_z，i分别将先行指标对称变化率C_w,i(t)和一致指标对称变化率C_z,i(t)进行标准化处理，以得到标准化变化率S_w,i(t)和S_z,i(t)，其中，In the above-mentioned embodiments of the present application, the results obtained after standardization and trend adjustment of the leading index symmetric rate of change C _w,i (t) and the consistent index symmetric rate of change C _z,i (t) are synthetically calculated to The steps of obtaining the preceding composite index and the consistent composite index include: obtaining the normalization factors A _w,i and A _z,i by the following formula:

t=2,3,...,n; use standardized factors A _w,i and A _z,i to respectively carry out the symmetrical change rate C _w,i (t) of the leading index and the symmetrical change rate C _z,i (t) of the consistent index Standardized processing to obtain the standardized rate of change S _w,i (t) and S _z,i (t), where,

$S_{w,i}\left(t\right)=\frac{C_{w,i}\left(t\right)}{A_{w,i}},$ $S_{z,i}\left(t\right)=\frac{C_{z,i}\left(t\right)}{A_{z,i}},$ t=2,3,...,n;

Perform average rate-of-change processing on the standardized rate of change S _w,i (t) and S _z,i (t) to obtain the standardized average rate of change V _w (t) of the leading index and the standardized average rate of change V _z ( t); according to the normalized average rate of change V _w (t) of the leading index and the normalized average rate of change V _z (t) of the consistent index, the composite calculation is performed to obtain the leading composite index I _w (t) and the consistent composite index I _z ( t), where, $I_w\left(t\right)=I_w(t-1)\&times;\frac{200+V_w\left(t\right)}{200-V_w\left(t\right)},$ $I_z\left(t\right)=I_z(t-1)\&times;\frac{200+V_z\left(t\right)}{200-V_z\left(t\right)},$ And I _w (1)=100, I _z (1)=100.

Specifically, according to the following formula, standardize the leading index symmetric rate of change C _w,i (t) and the consistent index symmetric rate of change C _z,i (t), so that the average absolute value is equal to 1, and the standardized factor A _{w, i} and _Az,i :

I _t , $A_{z,i}=\underset{t=2}{\overset{\mathrm{no}}{\mathrm{\&Sigma;}}}\frac{\left|C_{z,i}\left(t\right)\right|}{\mathrm{no}-1},$ t=2,3,...,n,

Then standardize C _w,i (t) and C _z,i (t) according to A _w,i and A _z, i respectively, and obtain the standardized rate of change S _w,i (t) and S _z,i (t):

t=2,3,…,n，之后根据标准化变化率S_w,i(t)和S_z,i(t)进行平均变化率处理，以获取先行指标的标准化平均变化率V_w(t)和一致指标的标准化平均变化率V_z(t)，并根据先行指标的标准化平均变化率V_w(t)和一致指标的标准化平均变化率V_z(t)进行合成计算，以获取先行合成指数I_w(t)和一致合成指数I_z(t)，具体地：I_w(1)=100，I_z(1)=100，则

t=2,3,...,n, and then carry out the average change rate processing according to the standardized change rate S _w,i (t) and S _z,i (t) to obtain the standardized average change rate V _w (t) of the leading indicator and the standardized average rate of change V _z (t) of the consistent index, and perform synthetic calculations based on the standardized average rate of change V _w (t) of the leading index and the standardized average rate of change V _z (t) of the consistent index to obtain the leading composite index I _w (t) and consistent synthetic index I _z (t), specifically: I _w (1)=100, I _z (1)=100, then

${\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&times;</span>\frac{\mathrm{<span\; class="notranslate">\; 200</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; V</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; 200</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; V</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; ,</span>$ ${\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&times;</span>\frac{\mathrm{<span\; class="notranslate">\; 200</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; V</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; 200</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; V</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; .</span>$

More specifically, after obtaining the leading composite index I _w (t) and the consistent composite index I _z (t), the power demand trend can be adjusted in the following way:

According to the following compound interest formula, calculate the respective average growth rate for each sequence of the consistent index group:

$r_i=(\sqrt[m_i]{C_{\mathrm{Li}}/C_{\mathrm{II}}}-1)\&times;100,$ i=1,2,...,k _z ,

和

分别是一致指标组第i个指标最先与最后循环的平均值，m_Ii与m_Li分别是一致指标组第i个指标最先与最后循环的月数，k₂是一致指标个数，m_i是最先循环的中心到最后循环的中心之间的月数。Among them, as shown in Figure 6, t is the month,

and

are the average values of the first and last cycles of the i-th index of the consistent index group, m _Ii and m _Li are the months of the first and last cycles of the i-th index of the consistent index group, k ₂ is the number of consistent indicators, m _i is the number of months between the center of the first cycle to the center of the last cycle.

之后对先行和一致指标的初始合成指数I_w(t)和I_z(t)分别用复利公式求出他们各自的平均增长率r′_w和r′_z：Then calculate the average growth rate G _r of the consistent index group and use it as the target trend:

Then use the compound interest formula to calculate their respective average growth rates r′ _w and r′ _z for the initial composite indices I _w (t) and I _z (t) of the leading and consistent indicators:

${\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>\sqrt[{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}]{{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; Lw</span>}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; Iw</span>}}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&times;</span>\mathrm{<span\; class="notranslate">\; 100</span>}<span\; class="notranslate">\; ,</span>$ ${\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>\sqrt[{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}]{{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; Lz</span>}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; Iz</span>}}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&times;</span>\mathrm{<span\; class="notranslate">\; 100</span>}<span\; class="notranslate">\; ,</span>$

in,

Then make trend adjustments for the standardized average rate of change V _w (t) and V _z (t) of the leading index group and the consistent index group:

V′ _w (t)=V _w (t)+(Gr-r′ _w ), V′ _z (t)=V _z (t)+(Gr-r′ _z ), t=2,3,…, n. Then calculate composite index according to the method in above-mentioned embodiment: make I' _w (1)=100, I' _z (1)=100, then

${{\mathrm{<span\; class="notranslate">\; I</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{{\mathrm{<span\; class="notranslate">\; I</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&times;</span>\frac{\mathrm{<span\; class="notranslate">\; 200</span>}<span\; class="notranslate">\; +</span>{{\mathrm{<span\; class="notranslate">\; V</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; 200</span>}<span\; class="notranslate">\; -</span>{{\mathrm{<span\; class="notranslate">\; V</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; ,</span>$ ${{\mathrm{<span\; class="notranslate">\; I</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{{\mathrm{<span\; class="notranslate">\; I</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&times;</span>\frac{\mathrm{<span\; class="notranslate">\; 200</span>}<span\; class="notranslate">\; +</span>{{\mathrm{<span\; class="notranslate">\; V</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; 200</span>}<span\; class="notranslate">\; -</span>{{\mathrm{<span\; class="notranslate">\; V</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; ,</span>$

Generate the leading composite index CI _w (t) and consistent composite index CI _z (t) with the base year as 100:

${\mathrm{<span\; class="notranslate">\; CI</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span>\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}<span\; class="notranslate">\; \&times;</span>\mathrm{<span\; class="notranslate">\; 100</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>$ ${\mathrm{<span\; class="notranslate">\; CI</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span>\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&times;</span>\mathrm{<span\; class="notranslate">\; 100</span>}<span\; class="notranslate">\; ,</span>$

和

分别是I′_w(t)和I′_z(t)在基准年份的平均值。in

and

are the mean values of I′ _w (t) and I′ _z (t) in the base year, respectively.

According to the above-mentioned embodiment of the present application, the average rate of change processing is performed on the standardized rate of change S _w,i (t) and S _z,i (t), so as to obtain the normalized average rate of change V _w (t) of the leading index and the consistent index The steps of the normalized average rate of change V _z (t) include:

The standardized rate of change S _w,i (t) of the leading index and the standardized rate of change S _z,i (t) of the consistent index are processed by the average rate of change respectively by the following formula to obtain the average rate of change R _w (t ) and the average rate of change R _z (t) of the consensus indicator:

$R_w\left(t\right)=\frac{\underset{i=1}{\overset{k_w}{\mathrm{\&Sigma;}}}{S}_{w,i}\left(t\right){\mathrm{\&lambda;}}_{w,i}}{\underset{i=1}{\overset{k_w}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_{w,i}},$ $R_z\left(t\right)=\frac{\underset{i=1}{\overset{k_z}{\mathrm{\&Sigma;}}}{S}_{z,i}\left(t\right){\mathrm{\&lambda;}}_{z,i}}{\underset{i=1}{\overset{k_z}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_{z,i}},$ Among them, λw _,i and λz _,i are the weights of the leading indicator and the i-th indicator of the consistent indicator respectively; the indicator standardization factor F _w is obtained by the following formula:

$f_w=\left[\underset{t=2}{\overset{\mathrm{no}}{\mathrm{\&Sigma;}}}\right|R_w\left(t\right)|/(\mathrm{no}-1)]/\left[\underset{t=2}{\overset{\mathrm{no}}{\mathrm{\&Sigma;}}}\right|R_z\left(t\right)|/(\mathrm{no}-1)];$ According to the standardization factor F _w of the index, the normalized average rate of change is processed to obtain the normalized average rate of change V _w (t) of the leading index and the normalized average rate of change V _z (t) of the consistent index, where V _w (t) = R _w (t)/F _w , V _z (t) = R _z (t).

Specifically, before obtaining the initial synthetic indices I _w (t) and I _z (t), the standardized rate of change S _w,i (t) of the leading index and the standardized rate of change S _z,i of the consistent index are respectively calculated by the following formula (t) Perform average rate of change processing to obtain the average rate of change R _w (t) of leading indicators and the average rate of change R _z (t) of consistent indicators:

$R_w\left(t\right)=\frac{\underset{i=1}{\overset{k_w}{\mathrm{\&Sigma;}}}{S}_{w,i}\left(t\right){\mathrm{\&lambda;}}_{w,i}}{\underset{i=1}{\overset{k_w}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_{w,i}},$ $R_z\left(t\right)=\frac{\underset{i=1}{\overset{k_z}{\mathrm{\&Sigma;}}}{S}_{z,i}\left(t\right){\mathrm{\&lambda;}}_{z,i}}{\underset{i=1}{\overset{k_Z}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_{z,i}},$ λ _w,i and λ _z,i are the weights of the i-th indicator of the leading and consistent indicator groups, respectively.

Then calculate the exponential normalization factor F _w according to the following formula:

${\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; [</span>\underset{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 2</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; |</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; [</span>\underset{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 2</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; |</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; ;</span>$

Finally, the normalized average rate of change is processed according to the index standardization factor F _w to obtain the standardized average rate of change V _w (t) of the leading index and the standardized average rate of change V _z (t) of the consistent index:

V _w (t)=R _w (t)/F _w , V _z (t)=R _z (t), t=2,3,…,n, where the amplitude of the average rate of change of the consistent index sequence is used to The purpose of adjusting the average rate of change of the leading index series and the lagging index series is to apply the two indexes as a coordinated system.

It should be noted that the steps shown in the flowcharts of the accompanying drawings may be performed in a computer system, such as a set of computer-executable instructions, and that although a logical order is shown in the flowcharts, in some cases, The steps shown or described may be performed in an order different than here.

From the above description, it can be seen that the present invention achieves the following technical effects: through the method and device for obtaining early warning parameters of electric power demand of the present application, after obtaining the trend item and period item in the original data sequence, by analyzing the data Sequence screening and calculation to obtain leading indicators and consistent indicators, and then use the index synthesis model to synthesize the above leading indicators and consistent indicators to obtain early warning parameters, and analyze the periodic fluctuations of power demand according to the early warning parameters, which solves the problem of power demand in the prior art. Forecasting technology obtains short-term periodic fluctuation distortion, and cannot obtain early warning parameters that meet power demand, which leads to the inability to formulate reasonable measures to deal with periodic fluctuations based on periodic power fluctuations. According to the effect of formulating reasonable and scientific countermeasures based on short-term cyclical fluctuations, the magnitude of cyclical fluctuations is slowed down, and the degree of damage caused by cyclical fluctuations to the power industry and economic development is reduced.

Obviously, those skilled in the art should understand that each module or each step of the above-mentioned present invention can be realized by a general-purpose computing device, and they can be concentrated on a single computing device, or distributed in a network formed by multiple computing devices Optionally, they can be implemented with program codes executable by a computing device, so that they can be stored in a storage device and executed by a computing device, or they can be made into individual integrated circuit modules, or they can be integrated into Multiple modules or steps are fabricated into a single integrated circuit module to realize. As such, the present invention is not limited to any specific combination of hardware and software.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and changes. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims (14)

1. A method for obtaining early warning parameters of electric power demand, characterized in that, comprising: Obtain the data series used to generate early warning indicators; Filtering the data series according to the adjustment parameters to obtain a data series containing trend items and period items; Calculating the trend index of the data sequence containing the trend item and the cycle item, and filtering the data sequence containing the trend item and cycle item according to the trend index to obtain the early warning index sequence, so The above-mentioned early warning index sequence is the data whose trend index is increasing in the data sequence; Extracting the power generation in the early warning index sequence as a benchmark index, and extracting indexes other than the power generation as selected indexes; According to the time difference analysis model and/or the K-L information amount model, the correlation calculation is performed on the early warning indicators, so as to obtain the correlation coefficient between each of the selected indicators and the benchmark index, and according to the correlation coefficient The selected indicators are screened to obtain leading indicators and consistent indicators; The leading index and the consistent index are synthesized according to the synthetic index model to obtain the leading synthetic index and the consistent synthetic index as early warning parameters. 2. method according to claim 1, is characterized in that, utilizes time-difference analysis model and/or K-L information amount model to carry out correlation calculation to described warning index, with each described selected index and described reference index The correlation coefficient between, and according to the correlation coefficient, the selected index is screened to obtain the steps of leading index and consistent index including: The correlation coefficient r _l between each selected index and the benchmark index is obtained according to the following formula: ${\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; =</span>\frac{\underset{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; -</span>\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; -</span>\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; the\; y</span>}}<span\; class="notranslate">\; )</span>}{\sqrt{\underset{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; -</span>\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; -</span>\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; the\; y</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}}<span\; class="notranslate">\; ,</span>$ Among them, l=0,±1,±2,…,±12, Y=(y ₁ ,y ₂ ,…,y _n ) is the benchmark index, X=(x ₁ ,x ₂ ,…,x _n ) is Selected indicators, l is the time difference, n _l is the number of all indicators, t=1,2,...,n is the number of months; Taking the selected index whose value of the time difference is within the first value range and the correlation coefficient r _l is greater than the first threshold as the leading index, and setting the value of the time difference within the second value range and The selected index whose correlation coefficient r _l is greater than the second threshold is used as the consistent index. 3. The method according to claim 2, characterized in that, according to the time difference analysis model and/or the K-L information volume model, the correlation calculation is carried out to the early warning index, and the selected index and the benchmark index are calculated according to each of the selected indexes. The correlation coefficient between, and according to the correlation coefficient, the selected index is screened to obtain the steps of leading index and consistent index including: The correlation coefficient r _l between each selected index and the benchmark index is obtained according to the following formula: ${\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; =</span>\frac{\underset{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; -</span>\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; -</span>\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; the\; y</span>}}<span\; class="notranslate">\; )</span>}{\sqrt{\underset{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; -</span>\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; -</span>\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; the\; y</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}}<span\; class="notranslate">\; ,</span>$ Among them, l=0,±1,±2,…,±12, Y=(y ₁ ,y ₂ ,…,y _n ) is the benchmark index, X=(x ₁ ,x ₂ ,…,x _n ) is Selected indicators, l is the time difference, n _l is the number of all indicators, t=1,2,...,n is the number of months; The time difference value is within the first value range and the selected index whose correlation coefficient r _l is greater than the first threshold is used as the initial index of the leading index, and the time difference value is in the second value range The selected index within the range and the correlation coefficient r _l greater than the second threshold is used as the initial index of the consistent index; Standardize the benchmark index, the initial index of the leading index, and the initial index of the consistent index to obtain a standard benchmark index sequence p _t and a sequence q _t of standard selected indexes, wherein the standard is selected Indicators include standard leading indicators and standard consistent indicators; The KL information amount k _l between each standard selected index and the standard benchmark index is obtained according to the following formula: k _l =∑p _t ln(p _t /q _t+1 ), among them, l=0,±1,…,±12,

t=1,2,..., n is the number of months, l is the time difference, n _l is the number of all indicators; The time difference value is within the third value range and the KL information amount k _l is less than the third threshold, and the standard selected index is used as the leading index, and the time difference value is within the fourth value range And the selected index whose KL information amount k _l is smaller than the fourth threshold is used as the consistent index. 4. according to the described method of claim 2 or 3, it is characterized in that, according to synthetic index model, described leading index and consistent index are synthesized, to obtain the step of leading synthetic index and consistent synthetic index as early warning parameter comprising: Symmetrical change processing is performed on the leading index and the consistent index respectively to obtain the symmetrical rate of change C _w,i (t) of the leading index and the symmetrical rate of change C _z,i (t) of the consistent index, Wherein, the leading index is subjected to symmetrical change processing by the following formula, so as to obtain the symmetrical change rate C _w,i (t) of the leading index: in,

is the i-th (i=1, 2, ..., k _w ) leading indicator, t=2, 3, ..., n, k _w is the number of leading indicators; Perform symmetrical change processing on the consistent index by the following formula to obtain the symmetrical rate of change C _z,i (t) of the consistent index:

in,

is the i (i=1, 2, ..., k _z ) consistent index, t = 2, 3, ..., n is the number of months, and k _z is the number of consistent indicators; Composite calculation is performed on the results obtained after standardization and trend adjustment of the leading index symmetrical rate of change C _w,i (t) and the consistent index symmetrical rate of change C _z,i (t), to obtain the leading synthetic index and consistent composite index. 5. The method according to claim 4, characterized in that, standardizing and trending the leading index symmetric rate of change C _w,i (t) and the consistent index symmetric rate of change C _z,i (t) The results obtained after adjustment are combined and calculated to obtain the preceding combined index and consistent combined index. The steps include: The normalization factors A _w,i and A _z,i are obtained by the following formula:

t=2,3,...,n; Using the normalization factors A _w,i and A _z,i to standardize the leading index symmetric rate of change C _w,i (t) and the consistent index symmetric rate of change C _z,i (t), respectively, to Get the standardized rate of change S _w,i (t) and S _z,i (t), where, $S_{w,i}\left(t\right)=\frac{C_{w,i}\left(t\right)}{A_{w,i}},$ $S_{z,i}\left(t\right)=\frac{C_{z,i}\left(t\right)}{A_{z,i}},$ t=2,3,...,n; Performing average rate of change processing on the standardized rate of change S _w,i (t) and S _z,i (t) to obtain the normalized average rate of change V _w (t) of the leading indicator and the normalized rate of the consistent indicator Average rate of change V _z (t); Composite calculation is performed according to the normalized average rate of change V _w (t) of the leading index and the normalized average rate of change V _z (t) of the consistent index to obtain the leading composite index I _w (t) and the consistent composite index I _z (t), where, $I_w\left(t\right)=I_w(t-1)\&times;\frac{200+V_w\left(t\right)}{200-V_{w}t},$ $I_z\left(t\right)=I_z(t-1)\&times;\frac{200+V_z\left(t\right)}{200-V_z\left(t\right)}.$ And I _w (1)=100, I _z (1)=100. 6. The method according to claim 5, characterized in that, the average rate of change process is performed on the normalized rate of change S _w,i (t) and S _z,i (t), to obtain the normalized rate of the leading index The steps of the average rate of change V _w (t) and the normalized average rate of change V _z (t) of the consistent index include: The standardized rate of change S _w,i (t) of the leading index and the standardized rate of change S _z,i (t) of the consistent index are respectively processed by the following formula to obtain the average rate of change of the leading index The rate of change R _w (t) and the average rate of change R _z (t) of the agreement index: $R_w\left(t\right)=\frac{\underset{i=1}{\overset{k_w}{\mathrm{\&Sigma;}}}{S}_{w,i}\left(t\right){\mathrm{\&lambda;}}_{w,i}}{\underset{i=1}{\overset{k_w}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_{w,i}},$ $R_z\left(t\right)=\frac{\underset{i=1}{\overset{k_z}{\mathrm{\&Sigma;}}}{S}_{z,i}\left(t\right){\mathrm{\&lambda;}}_{z,i}}{\underset{i=1}{\overset{k_z}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_{z,i}},$ Among them, λw _,i and λz _,i are the weights of the i-th index of the leading index and the consistent index respectively; Obtain the index normalization factor F _w by the following formula: ${\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; [</span>\underset{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 2</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; |</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; [</span>\underset{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 2</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; |</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; ;</span>$ According to the standardization factor F _w of the index, the normalized average rate of change is processed to obtain the normalized average rate of change V _w (t) of the leading index and the normalized average rate of change V _z (t) of the consistent index, wherein V _w (t)=R _w (t)/F _w , V _z (t)=R _z (t). 7. The method according to claim 1, characterized in that, after obtaining the data sequence for generating early warning indicators, the method further comprises: Perform preprocessing on the data in the data sequence, the preprocessing includes: filling missing data processing, correcting noise data processing, data smoothing processing and data normalization processing. 8. A device for obtaining early warning parameters of power demand, characterized in that it comprises: The first acquisition module is used to acquire the data sequence used to generate early warning indicators; The first processing module is used to filter the data series according to the adjustment parameters, so as to obtain the data series including trend items and period items; The first calculation module is used to calculate the trend index of the data sequence containing the trend item and the periodic item, and filter the data sequence containing the trend item and the periodic item according to the trend index, To obtain the early warning index sequence, the early warning index sequence is the data whose trend index is increasing in the data sequence; The first extraction module is used to extract the power generation in the early warning index sequence as a benchmark index, and extract indexes other than the power generation as selected indexes; The second calculation module is used to calculate the correlation of the early warning indicators according to the time difference analysis model and/or the K-L information model, so as to obtain the correlation coefficient between each of the selected indicators and the benchmark indicator, and Screening the selected indicators according to the correlation coefficient to obtain leading indicators and consistent indicators; The second processing module is used for synthesizing the leading index and the consistent index according to the synthetic index model, so as to obtain the leading synthetic index and the consistent synthetic index as early warning parameters. 9. The device according to claim 8, wherein the second calculation module comprises: The first sub-calculation module is used to obtain the correlation coefficient r _l between each of the selected indicators and the benchmark indicator according to the following formula: Among them, l=0,±1,±2,…,±12, Y=(y ₁ ,y ₂ ,…,y _n ) is the benchmark index, X=(x ₁ ,x ₂ ,…,x _n ) is Selected indicators, l is the time difference, n _l is the number of all indicators, t=1,2,...,n is the number of months; The first sub-processing module is configured to use the selected index whose value of the time difference is within the first value range and the correlation coefficient r _l is greater than the first threshold as the leading index, and take the value of the time difference The selected index within the second value range and the correlation coefficient r _l greater than the second threshold is used as the consistent index. 10. The device according to claim 9, wherein the second calculation module comprises: The second sub-calculation module is used to obtain the correlation coefficient r _l between each selected index and the benchmark index according to the following formula:

Among them, l=0,±1,±2,…,±12, Y=(y ₁ ,y ₂ ,…,y _n ) is the benchmark index, X=(x ₁ ,x ₂ ,…,x _n ) is Selected indicators, l is the time difference, n _l is the number of all indicators, t=1,2,...,n is the number of months; The second sub-processing module is used to use the selected index whose value of the time difference is within the first value range and the correlation coefficient r _l is greater than the first threshold as the initial index of the leading index, and the The time difference value is within the second value range and the selected index whose correlation coefficient r _l is greater than the second threshold is used as the initial index of the consistent index; The third sub-processing module is used to standardize the benchmark index, the initial index of the leading index, and the initial index of the consistent index, so as to obtain the standard benchmark index sequence p _t and the standard selected index sequence q _t , wherein, the standard selected indicators include standard leading indicators and standard consistent indicators; The third sub-calculation module is used to obtain the KL information amount k _l between each standard selected index and the standard benchmark index according to the following formula: k _l =∑p _t ln(p _t /q _t+1 ), where l=0，±1，…,±12，

t=1, 2, ..., n is the number of months, l is the time difference, n _l is the number of all indicators; The fourth sub-processing module is used to use the time difference value within the third value range and the KL information amount k _l is less than the third threshold as the standard selected index as the leading index, and take the time difference A selected indicator whose value is within a fourth value range and whose KL information amount k _l is smaller than a fourth threshold is used as the consistent indicator. 11. The device according to claim 9 or 10, wherein the second processing module comprises: The fifth sub-processing module is used to perform symmetrical change processing on the leading index and the consistent index respectively, so as to obtain the symmetrical rate of change of the leading index C _w,i (t) and the symmetrical rate of change of the consistent index C _z,i (t ), the fifth sub-processing module includes: The fourth sub-calculation module is used to perform symmetrical change processing on the leading index through the following formula, so as to obtain the symmetrical rate of change C _w,i (t) of the leading index:

in,

is the i-th (i=1, 2, ..., k _w ) leading indicator, t=2, 3, ..., n is the number of months, and k _w is the number of leading indicators; The fifth sub-calculation module is used to perform symmetrical change processing on the consistent index through the following formula, so as to obtain the symmetrical change rate C _z,i (t) of the consistent index:

in,

is the i-th (i=1, 2, ..., k _z ) consistent index, t=2, 3, ..., n, k _z is the number of consistent indicators; The sixth sub-processing module is used to synthesize the results obtained after standardization and trend adjustment of the leading indicator symmetric rate of change _Cw,i (t) and the consistent indicator symmetric rate of change _Cz,i (t) Calculated to obtain the preceding composite index and the consistent composite index. 12. The device according to claim 11, wherein the sixth sub-processing module comprises: The sixth sub-calculation module is used to obtain the normalization factors A _{w, i} and A _{z, i} through the following formula:

$A_{z,i}=\underset{t=2}{\overset{\mathrm{no}}{\mathrm{\&Sigma;}}}\frac{\left|C_{z,i}\left(t\right)\right|}{\mathrm{no}-1},$ t=2,3,...,n; The seventh sub-processing module is used to use the normalization factors A _w,i and A _z,i to respectively convert the symmetrical rate of change C _w,i (t) of the leading index and the symmetrical rate of change C _z,i of the consistent index (t) carry out standardized processing to obtain the standardized rate of change S _w,i (t) and S _z,i (t), where, $S_{w,i}\left(t\right)=\frac{C_{w,i}\left(t\right)}{A_{w,i}},$ $S_{z,i}\left(t\right)=\frac{C_{z,i}\left(t\right)}{A_{z,i}},$ t=2,3,...,n; The eighth sub-processing module is used to perform average rate-of-change processing on the standardized rate-of-change S _w,i (t) and S _z,i (t), so as to obtain the normalized average rate-of-change V _w (t ) and the normalized average rate of change V _z (t) of the consistent index; The seventh sub-calculation module is used to perform composite calculation according to the normalized average rate of change V _w (t) of the leading index and the normalized average rate of change V _z (t) of the consistent index, so as to obtain the leading composite index I _w (t) and consistent composite index I _z (t), where, $I_w\left(t\right)=I_w(t-1)\&times;\frac{200+V_w\left(t\right)}{200-V_{w}t},$ $I_z\left(t\right)=I_z(t-1)\&times;\frac{200+V_z\left(t\right)}{200-V_z\left(t\right)}.$ And I _w (1)=100, I _z (1)=100. 13. The device according to claim 12, wherein the eighth sub-processing module comprises: The ninth sub-processing module is used to process the average rate of change of the standardized rate of change S _w,i (t) of the leading index and the standardized rate of change S _z,i (t) of the consistent index respectively by the following formula, To obtain the average rate of change R _w (t) of the leading indicator and the average rate of change R _z (t) of the consistent indicator: $R_w\left(t\right)=\frac{\underset{i=1}{\overset{k_w}{\mathrm{\&Sigma;}}}{S}_{w,i}\left(t\right){\mathrm{\&lambda;}}_{w,i}}{\underset{i=1}{\overset{k_w}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_{w,i}},$ $R_z\left(t\right)=\frac{\underset{i=1}{\overset{k_z}{\mathrm{\&Sigma;}}}{S}_{z,i}\left(t\right){\mathrm{\&lambda;}}_{z,i}}{\underset{i=1}{\overset{k_z}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_{z,i}},$ Among them, λ _w,i and λ _z,i are the weights of the i-th index of the leading index and the consistent index respectively; The eighth sub-calculation module is used to obtain the index normalization factor F _w through the following formula: ${\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; [</span>\underset{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 2</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; |</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; [</span>\underset{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 2</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; |</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; ;</span>$ The ninth sub-computing module is used to process the normalized average rate of change according to the index normalization factor _Fw , so as to obtain the normalized average rate of change _Vw (t) of the leading index and the normalized average rate of change V of the consistent index _z (t), wherein, V _w (t) = R _w (t) / F _w , V _z (t) = R _z (t). 14. The device according to claim 8, wherein after executing the obtaining module, the device further comprises: The tenth sub-processing module is used for preprocessing the data in the data sequence, and the preprocessing includes: filling missing data processing, correcting noise data processing, data smoothing processing and data normalization processing.

Images