CN110009165A - Prediction method and system of historical electricity consumption based on multi-address fence area - Google Patents

Abstract

The invention relates to a prediction system based on the historical power consumption of multi-address fence areas, including a database, a memory and a server connected by communication, wherein the historical power consumption data of N address fence areas are stored in the database E={E ₁ ,E ₂ ,...,E _N }, where the historical power consumption data of the i-th address fence area E _i =(e _i(‑ti) ,e _i(‑ti+1) ,...,e _i(-1) ), the value of i ranges from 1 to N; an application program is stored in the memory, and the following steps are implemented when the application program is executed by the server: Step S100, according to historical power consumption data E predicts the electricity consumption of each geo-fence area in the N address fence areas {e ₁₀ ,e ₂₀ ,...,e _N0 }, where e _i0 is the predicted electricity consumption of the ith address fence area; step S200, determine the predicted power consumption of the N address fence areas .

Description

Prediction method and system of historical electricity consumption based on multi-address fence area

technical field

The invention relates to the field of big data analysis and electricity consumption prediction, in particular to a prediction system based on historical electricity consumption in a multi-address fence area.

Background technique

The forecast of electricity consumption in medium and large regions (such as development zones, prefecture-level cities, provincial regions, large regions including multiple provinces, and the whole country) has always been a hot topic in the research of the power industry. The complexity of electricity consumption forecasting in a specific fenced area, for example, the impact of changes in electricity consumption in a fenced area that will increase or decrease in a short period of time on other fenced areas. Data technology to predict electricity consumption in medium-sized fenced areas is a technical problem that needs to be solved urgently

SUMMARY OF THE INVENTION

In view of this, the purpose of the present invention is to provide a prediction method based on the historical power consumption of a multi-address fence area, which can predict the power consumption of a medium-sized fence area according to the historical power consumption and combined with big data technology.

To achieve the above object, the present invention adopts the following technical solutions:

A prediction method based on historical power consumption in a multi-address fenced area provides a prediction system including a database, a memory and a server connected by communication, and specifically includes the following steps:

Include the following steps:

Step S1: collect historical power consumption data _{E =} {E ₁ , E ₂ , . _i(-ti) ,e _i( -ti+1),...,e _i( -1)), the value of i ranges from 1 to N;

Step _S2 : _Predict the electricity consumption {e ₁₀ , e ₂₀ , . The electricity consumption of the address fence area;

Step S3: According to the electricity consumption of each geographic fence area, obtain the predicted electricity consumption of N address fence areas

Step S4: Present the E0 on the display device or send it to the user's mobile terminal for display.

Further, the prediction system also includes a data cache, and the data cache stores a regional correlation matrix where p _ij is the correlation parameter of the jth address fence area to the electricity consumption of the i th address fence area.

Further, the predicted electricity consumption of the i-th address fence area

Among them, f(E _i ) is the electricity consumption of the ith address fence area predicted according to E _i , e _j(-1) and e _j(-2) are the current time period and the previous time period in the j th address fence area The actual electricity consumption for a time period; e _j(-1) and e _j(-2) are the actual power consumption in the current time period and the previous time period in the jth address fence area, and D1 is a preset first threshold.

Further, the D1=f(e _j(-1) , e _j(-2) ), that is, D1 is a function of e _j(-1) and e _j(-2) .

Further, the said Δe _ij(-1) is the difference between the power consumption of the i-th and j-th address fence areas in the current time period, and minΔe _ij and maxΔe _ij are the historical power consumption of the i-th and j-th address fence areas in the same time period respectively. The difference between the minimum and maximum value of historical electricity consumption, η _ij ∈(0,1).

Further, the prediction system further includes a data cache, where N region association sequences Q ₁ , Q ₂ ,...,Q _N are stored in the data cache, wherein, ID _k is the unique identifier of the address fence area that can be associated with the predicted electricity consumption of the _ith address fence area Qi, It is the correlation parameter of the electricity consumption of the ith address fence area for the address fence area of ID _k .

Further, the said where f(E _i ) is the electricity consumption of the ith address fence area predicted according to E _i , e _IDk(-1) and e _IDk(-2) are the actual power consumption in the current time period and the previous time period in the address fence area of the address fence area that identifies ID _k , and D1 is a preset first threshold.

Compared with the prior art, the present invention has the following beneficial effects:

The present invention can predict the power consumption of a medium-sized fence area according to the historical power consumption and combined with the big data technology, and the prediction accuracy is high.

Description of drawings

Fig. 1 is the flow chart of the method of the present invention.

Detailed ways

The present invention will be further described below with reference to the accompanying drawings and embodiments.

Referring to FIG. 1, the present invention provides a prediction method based on the historical power consumption of a multi-address fence area, and provides a prediction system including a database, a memory, a server and a data cache connected by communication. The database and the data cache can be implemented as software, or can be Implemented as hardware, any device or software module in the prior art that can implement corresponding functions can be applied to the present invention. In the present invention, communication connections may include bus connections within devices or wired and/or wireless connections between devices.

In this embodiment, the following steps are included:

Step S1: collect historical power consumption data _{E =} {E ₁ , E ₂ , . _i(-ti) ,e _i(-ti+1) ,...,e _i(-1) ), the value of i ranges from 1 to N;

In this embodiment, an application program is stored in the memory, and the following steps are implemented when the application program is executed by the server:

Step _S2 : _Predict the electricity consumption {e ₁₀ , e ₂₀ , . The electricity consumption of the address fence area;

Step S3: According to the electricity consumption of each geographic fence area, obtain the predicted electricity consumption of N address fence areas

Step S4: Present the E0 on the display device or send it to the user's mobile terminal for display.

In this embodiment, the prediction system further includes a data cache, and the data cache stores a regional correlation matrix where p _ij is the associated parameter of the jth address fence area to the power consumption of the i th address fence area. preferably, Among them, Δe _{ij (-1)} is the difference between the electricity consumption of the ith and j th address fence areas in the current time period, and minΔe _ij and maxΔe _ij are the same time in the history of the ith and j th address fence areas, respectively The difference between the minimum value and the maximum value of the historical power consumption of a period, η _ij is a resolution coefficient, η _ij ∈(0,1), preferably 0.5.

In this embodiment, the predicted electricity consumption of the i-th address fence area

Among them, f(E _i ) is the electricity consumption of the ith address fence area predicted according to E _i , e _j(-1) and e _j(-2) are the current time period and the previous time period in the j th address fence area The actual electricity consumption for a time period; e _j(-1) and e _j(-2) are the actual power consumption in the current time period and the previous time period in the jth address fence area, and D1 is a preset first threshold. The D1=f(e _j(-1) , e _j(-2) ), that is, D1 is a function of e _j(-1) and e _j(-2) . preferably, The value of l is about 3.41-3.67.

In another embodiment, the data cache stores N region association sequences Q ₁ , Q ₂ , . . . , Q _N , wherein, ID _k is the unique identifier of the address fence area that can be associated with the predicted electricity consumption of the _ith address fence area Qi, It is the correlation parameter of the electricity consumption of the ith address fence area for the address fence area of ID _k . said where f(E _i ) is the electricity consumption of the ith address fence area predicted according to E _i , e _IDk(-1) and e _IDk(-2) are the actual power consumption of the current time period and the previous time period in the address fence area of the address fence area that identifies ID _k , and D1 is the preset first threshold. In this embodiment, by using the correlation sequence instead of the correlation matrix in the previous embodiment, the increase or decrease of the number of correlation parameters is more convenient to handle; at the same time, in this embodiment, only the information related to the current predicted fence area that is related to the useful power is stored. The associated parameters are not considered, which makes the prediction of electricity consumption more convenient and efficient.

Step S200, determine the predicted power consumption of the N address fence areas

In step S300, the E0 is presented on the display device, or sent to the user's terminal device. According to the set terminal equipment, including computer equipment, mobile terminal equipment, etc., it can be sent through SMS, email, WeChat, etc.

The above descriptions are only preferred embodiments of the present invention, and all equivalent changes and modifications made according to the scope of the patent application of the present invention shall fall within the scope of the present invention.

Claims (7)

1. a kind of prediction technique based on multiaddress fence area history electricity consumption, which is characterized in that provide a forecasting system packet The database, memory and server of communication connection are included, specifically includes the following steps:

Step S1: the history electricity consumption data E={ E of N number of address fence area is acquired₁,E₂,...,E_N, wherein i-th of address The history electricity consumption data E of fence area_i=(e_i(-ti),e_i(-ti+1),...,e_i(-1)), the value range of i is from 1 to N；

Step S2: the electricity consumption of each geography fence region in N number of address fence area is predicted according to history electricity consumption data E {e₁₀,e₂₀,...,e_N0, wherein e_i0For the electricity consumption of i-th of address fence area of prediction；

Step S3: according to the electricity consumption of each geography fence region, the prediction electricity consumption of N number of address fence area is obtained

Step S4: the E0 mobile terminal for being presented on the display apparatus or being sent to user is shown.

2. the prediction technique according to claim 1 based on multiaddress fence area history electricity consumption, it is characterised in that: institute Stating forecasting system further includes data buffer storage, and region incidence matrix is stored in the data buffer storageWherein p_ijIt is j-th of address fence area to the pass of i-th of address fence area electricity consumption Join parameter.

3. the prediction technique according to claim 2 based on multiaddress fence area history electricity consumption, it is characterised in that: institute State the electricity consumption of i-th of address fence area of prediction

Wherein, f (E_i) it is according to E_iThe electricity consumption of i-th of address fence area of prediction, e_j(-1)And e_j(-2)It is enclosed for j-th of address The practical electricity consumption of current time period and a upper time cycle in column region；e_j(-1)And e_j(-2)For current time period in the fence area of j-th of address and The practical electricity consumption of a upper time cycle, D1 are preset first threshold.

4. the prediction technique according to claim 3 based on multiaddress fence area history electricity consumption, it is characterised in that: institute State D1=f (e_j(-1),e_j(-2)), i.e. D1 is e_j(-1)With e_j(-2)Function.

5. the prediction technique according to claim 3 based on multiaddress fence area history electricity consumption, it is characterised in that: institute It states

Δe_ij(-1)It is used for i-th in current time period and j-th of address fence area The difference of electricity, min Δ e_ijWith max Δ e_ijRespectively i-th and j-th of address fence area are in history the same as a period of time The minimum value and maximum value of the difference of history electricity consumption, η_ij∈(0,1)。

6. the prediction technique according to claim 1 based on multiaddress fence area history electricity consumption, it is characterised in that: institute Stating forecasting system further includes data buffer storage, and N number of region relating sequence Q is stored in data buffer storage₁,Q₂,...,Q_N, whereinID_kFor can be to i-th of address fence area Q_iPrediction electricity consumption generate pass The unique identification of the address fence area of connection,To identify ID_kAddress fence area to i-th of address fence area electricity consumption Relevant parameter.

7. the prediction technique according to claim 8 based on multiaddress fence area history electricity consumption, it is characterised in that: institute It statesWherein f (E_i) it is according to E_iThe electricity consumption of i-th of address fence area of prediction, WithTo identify ID_kAddress fence area address The practical electricity consumption of current time period and a upper time cycle in fence area, D1 are preset first threshold.