CN105184384A - Model for analyzing circulation characteristic factors affecting fog days and predicting fog days - Google Patents

Abstract

The invention discloses a factor analysis of circulation characteristic quantities affecting foggy days and a foggy day forecasting model. The meteorological observation data is used to analyze the relevant data of foggy days in different regions, and the correlation calculation formula of foggy days is used to find out the relationship between each circulation factor and each zone. According to the relationship between foggy days, the circulation characteristic factors that are better related to foggy days are screened out, and the main circulation characteristic factors that affect foggy days in different regions and in different months are obtained. Using the monthly average height field data of 500 hPa and the correlation analysis of foggy days, a monthly forecast model of foggy days in different regions is given by stepwise linear regression method. The present invention can output prediction results with a numerical model, and the prediction is carried out on the spot, and the model is suitable for heavy fog and haze weather prediction with visibility less than 1000m. Relevant units in the power sector can prepare various preventive work and emergency response plans in advance according to the forecast results, which plays an important role in effectively responding to pollution flashover accidents and hidden dangers in foggy weather.

Description

A Factor Analysis of Circulation Characteristic Quantities Affecting Foggy Days and a Foggy Day Prediction Model

technical field

The invention relates to a factor analysis of circulation characteristic quantities affecting foggy days and a foggy day prediction model, which belongs to the field of analysis and prediction of power grid operation environment.

Background technique

With the development of industry in recent years, severe haze has the characteristics of wide area, frequent occurrence and long duration of single haze. The wet and polluted environment of severe smog weather makes the transmission and transformation equipment of the power system have serious hidden dangers of pollution flashover, which poses a serious threat to the safe and stable operation of power transmission and transformation equipment. Since the formation of haze weather is not only affected by meteorological conditions, but also related to the increase of air pollutant emissions, and the monitoring data of haze is limited, it is very difficult to accurately predict the occurrence of haze weather.

Haze weather prediction is the basis for all sectors of society to reasonably and adequately deal with the hazards of haze weather. Judging from the research status at home and abroad, research on haze weather has achieved certain results, but in general, more research is Concentrate on the causes of haze weather, component analysis, meteorological feature analysis, and preventive and control measures. For the correlation between heavy haze weather (foggy days) and circulation characteristic factors, 500 hPa monthly average height field, and the number of heavy haze days There are few related researches on the forecasting method, and there is a lack of an effective method for forecasting the number of heavy haze days. Accurate and effective heavy haze weather prediction models play an important role in the power system's reasonable response to pollution flashover accidents in heavy haze weather. Insulator pollution is a large-scale regional problem. Once a pollution flashover accident occurs, it will affect a large area. Long duration is one of the most serious problems for the safe operation of dangerous power systems. Effective meteorological prediction methods for heavy fog and haze weather have important guiding significance for the prevention and control of insulator failures and hidden dangers.

Contents of the invention

The technical problem to be solved by the present invention is to provide a factor analysis of circulation characteristic quantities affecting foggy days and a foggy day prediction model, and the prediction model is suitable for heavy fog and haze weather.

In order to solve the above-mentioned technical problems, the factor analysis and foggy day forecasting model of the circulation feature quantity factor analysis and foggy day that the present invention provides, comprise the following steps:

(1) Collect, collate, and statistically analyze the 30-year historical meteorological data and circulation data in the provincial area, as well as the monthly foggy days from October to March of the following year (heavy haze period). Calculate the distribution characteristics of monthly foggy days, and divide the regions according to the number of foggy days and regional geographical environment conditions (such as: mountainous areas, piedmont plains, and plains);

(2) Calculate the correlation coefficients between the 74 circulation feature quantities released by the National Climate Center's Climate System Diagnosis and Prediction Office and the number of monthly foggy days according to the foggy day correlation calculation formula. The foggy day correlation calculation formula is:

${\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; no</span>}\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; \&CenterDot;</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}{\sqrt{\mathrm{<span\; class="notranslate">\; no</span>}\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span>{<span\; class="notranslate">\; (</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; \&CenterDot;</span>\sqrt{\mathrm{<span\; class="notranslate">\; no</span>}\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span>{<span\; class="notranslate">\; (</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}}$

In the formula, R _X is the correlation coefficient between the characteristic quantity factor of circulation and the number of fog days in a month, x _i is the value of the characteristic quantity factor of circulation, i is the serial number of the year, n is the number of historical years, and d _i is the number of foggy days in a certain month;

(3) From the calculation results of step (2), select the circulation characteristic factors with strong correlation in each month;

(4) Use NCEP/NCAR500hPa to reanalyze the height field data, and statistically analyze the 500hpa key area factors that affect the foggy days of each month;

(5) Using the stepwise regression analysis method to establish the forecast model of foggy days in each month.

Further, the fog days are heavy haze days with visibility less than 1000m. According to the distribution characteristics of fog days and geographical environment factors, different regions are divided, regional statistical analysis is carried out, and a regional prediction model is established.

Preferably, the grid point spacing of NCEP/NCAR 500hPa reanalysis height field data is 2.5°×2.5°, and the number of grid points is 144×37; in the process of mathematical statistical analysis, the 500 hPa monthly level height field and the predicted area are calculated grid by grid. For the linear correlation coefficient of foggy days in each month, the area with an area larger than 4×4 grid points is taken as the key area factor for foggy day prediction; in order to eliminate the error of a single grid point, the average value of the regional grid point value is used as the key area factor.

The present invention adopts the stepwise regression method to establish a prediction model, and provides the significance check (F check) and the root mean square error of the prediction model at the same time; the numerical model can be used to output the prediction result during the prediction, and the prediction is carried out on the spot.

The present invention carries out the research of macro aspect from the climatic law that fog (heavy haze) forms, utilizes historical meteorological data, 500 hPa monthly average height field data, carries out the correlation analysis of 500 hPa monthly average height field and fog day, through The stepwise linear regression method is used to form a month-by-month prediction model of foggy days in different regions. When the heavy foggy weather comes in autumn and winter (October to March of the next year), the monthly heavy foggy days are predicted.

The beneficial effects of adopting the above-mentioned technical solution are: the use of this prediction model realizes the real-time prediction of the number of heavy haze days in autumn and winter, and can effectively improve and improve the ability to predict and respond to such difficult and frequent disaster weather. As far as the power system is concerned, it is of great significance for the power grid to effectively deal with the pollution flashover accidents of power transmission and transformation equipment under heavy smog weather. For the provincial power grid area, according to the prediction conclusions of the number of heavy smog days in autumn and winter, prepare the corresponding power grid emergency response plan in advance to prevent large-scale power outages caused by grid pollution flashover accidents, reduce the losses caused by pollution flashover accidents, and ensure that the power grid security and stability has important guiding significance.

Description of drawings

Fig. 1 is a flowchart of the present invention.

Detailed ways

Below is an embodiment of the circulation feature factor analysis and the foggy day prediction model model analysis of the influence of the southern Hebei grid region on the foggy day with the inventive method, referring to accompanying drawing 1, and its concrete method is:

(1) First collect, sort out, and statistically analyze the number of foggy days in each month from October 1981 to March 2013 at each meteorological station in the Southern Hebei Power Grid area, the 74 circulation feature data released by the Climate System Diagnosis and Prediction Office of the National Climate Center, and the northern hemisphere NCEP/NCAR500hPa reanalysis height field data. According to the distribution characteristics of foggy days and geographical environment conditions, the southern Hebei power grid area is divided into three areas: mountainous area, piedmont plain area and eastern plain area.

(2) Calculate the linear correlation coefficient between the foggy days and the 74 circulation characteristic quantities of the three subregions from October to March in the southern Hebei power grid area. The absolute value of the correlation coefficient is greater than 0.344 (significant level α = 0.05) as the foggy day forecast factor. Tables 1 to 6 show the circulation characteristic factors affecting foggy days in each month.

Table 1. The characteristic quantity factors of the foggy day circulation in October

Table 2.11 Foggy Day Circulation Characteristic Factors

Table 3. The characteristic quantity factors of the foggy day circulation in December

Table 4. The characteristic quantity factors of the foggy day circulation in November

Table 5. The characteristic quantity factors of the foggy day circulation in February

Table 6. The characteristic quantity factors of the foggy day circulation in March

(3) Calculate the linear correlation coefficient of the 500 hPa monthly mean height field in the northern hemisphere and the foggy days in the three subregions (mountainous area, Shannan plain, and eastern plain) from October to March in the southern power grid area of Hebei Province on a grid-by-grid basis. The correlation coefficient is absolute The value is greater than 0.344, (significant level а=0.05), and the area with an area greater than 4×4 grid points is taken as the key area factor for foggy day prediction. In order to eliminate a single grid point error, the average value of the area grid point value is used as the key area factor here. Tables 7 to 12 give the number, correlation coefficient, center position and number of grid points of the precipitation key area factors in different regions in each month. Since there may be a good correlation between the factors in the key area, the method of stepwise regression is used to establish the prediction equation, and the significance test (F test) and root mean square error of the equation are given. Table 13 shows the month-by-month prediction model for the number of foggy days in different regions of the power grid in southern Hebei from October to March.

Table 7 Number of key regions and regional characteristic values affecting foggy days in October

Table 8. Factors of key areas of foggy days in November

Table 9 Key Area Factors of Foggy Days in December

Table 101 Foggy Day Key Area Factors

Table 11 Key area factors of foggy days in December

Table 12 Key area factors of foggy days in March

Table 13 Monthly prediction equation of monthly foggy days in different regions of southern Hebei power grid from October to March

With the method of the present invention, we can predict the number of heavy haze days per month during the frequent occurrence of haze in autumn and winter every year. For the provincial power grid area, the monthly live forecast of the number of smog days plays an important role in the relevant departments of the power system to reasonably respond to the occurrence of pollution flashover accidents under heavy smog and haze weather. Insulator pollution is a large-scale regional problem. Once it occurs Pollution flashover accidents have a large impact and a long duration, which is one of the most serious problems in the safe operation of dangerous power systems. An effective heavy haze weather prediction model has important guiding significance for the prevention and control of insulator failures and hidden dangers.

The present invention utilizes 74 items of circulation characteristic data and surface meteorological observation data publicly released by the Climate System Diagnosis and Prediction Office of the National Climate Center, and analyzes the relevant data of foggy days in different regions, and uses the correlation calculation formula of foggy days to find out the relationship between each circulation factor and each According to the relationship between foggy days in different regions, the circulation characteristic factors that are better related to foggy days are screened out, and the main circulation characteristic factors that affect foggy days in different regions and in different months are obtained. The monthly average height field data of 500 hPa is used for correlation analysis with foggy days, and the monthly forecast model of foggy days in different regions is given through stepwise linear regression method. The meteorological related data mentioned in this technical plan are all public information. The present invention can output prediction results with a numerical model, and the prediction is carried out on the spot, and the model is suitable for heavy fog and haze weather prediction with visibility less than 1000m. Relevant units in the power sector can prepare various preventive work and emergency response plans in advance according to the forecast results, which plays an important role in effectively responding to pollution flashover accidents and hidden dangers in foggy weather.

Claims (4)

1. affect mist day circulation characteristic factorial analysis and mist day a forecast model, it is characterized in that: it comprises the following steps:

(1) collect, arrange, the Historical Meteorological Information of 30 years, in March in October to the next year each mist day moon of circulation data and each meteorological site statistics in statistical study provincial region; Statistics moon mist day distribution characteristics, and according to mist number of days and regional geography environmental baseline, zoning: mountain area, pediment plain, region of no relief;

(2) according to mist day correlation calculations formulae discovery go out 74 circulation characteristic issuing National Climate center weather system diagnostics prediction room and the moon mist number of days related coefficient, mist day correlation calculations formula be:

$R_x=\frac{n\underset{i=1}{\overset{n}{\&Sigma;}}{x}_i{d}_i-\underset{i=1}{\overset{n}{\&Sigma;}}{x}_i\&CenterDot;\underset{i=1}{\overset{n}{\&Sigma;}}{d}_i}{\sqrt{n\underset{i=1}{\overset{n}{\&Sigma;}}{x}_i^2-{\left(\underset{i=1}{\overset{n}{\&Sigma;}}{x}_i\right)}^2}\&CenterDot;\sqrt{n\underset{i=1}{\overset{n}{\&Sigma;}}{d}_i^2-{\left(\underset{i=1}{\overset{n}{\&Sigma;}}{d}_i\right)}^2}}$

R in formula _xfor the circulation characteristic factor and the moon mist number of days related coefficient, x _ifor the value of the circulation characteristic factor, i is time sequence number, and n is historical years quantity, d _ifor the mist number of days of certain month;

(3) from the result of calculation of step (2), filter out the stronger circulation characteristic factor of each month correlativity;

(4) NCEP/NCAR500hPa is utilized to analyze height field data again, the 500hpa key area factor of the statistical study impact each mist day moon;

(5) Stepwise Regression Method is adopted to set up each mist moon, forecast model day.

2. according to claim 1 a kind of affect mist day circulation characteristic factorial analysis and mist day forecast model, it is characterized in that, described mist day is the heavy haze weather that visibility is less than 1000m.

3. according to claim 1 a kind of affect mist day circulation characteristic factorial analysis and mist day forecast model, it is characterized in that, the lattice point spacing that NCEP/NCAR500hPa analyzes height field data is again 2.5 ° × 2.5 °, and lattice point number is 144 × 37; In Mathematical Statistics Analysis process, by the grid computing 50,000 handkerchief moon flat height field and the linearly dependent coefficient of the estimation range each mist day moon, get the key area factor that region that region area is greater than 4 × 4 lattice points was predicted as mist day; For eliminating single lattice point error, using the mean value of region lattice point numerical value as the key area factor.

4. according to claim 1 a kind of affect mist day circulation characteristic factorial analysis and mist day forecast model, it is characterized in that, adopt the method establishment forecast model of successive Regression, provide checking validity and the root-mean-square error of forecast model simultaneously; By the numerical model prediction of output result during prediction.