CN-120952223-B - Multi-factor-associated method and system for pre-controlling power load of winter in rural area

CN120952223BCN 120952223 BCN120952223 BCN 120952223BCN-120952223-B

Abstract

The invention relates to the technical field of intelligent power grids, and discloses a multi-factor-associated rural power load pre-control method and system for winter, wherein the method comprises the steps of converting pre-acquired historical power consumption data to obtain power load data, and identifying the power load data to obtain a power consumption problem; analyzing the electricity utilization problem to obtain influence factors, analyzing the influence factors to obtain a strong correlation factor data matrix, analyzing the strong correlation factor data matrix to obtain a correlation influence coefficient, performing iterative training on a power load prediction model to obtain a trained power load prediction model, inputting current electricity utilization data obtained in advance into the trained power load prediction model to obtain load prediction data, dividing a power grid to obtain red-yellow-green classification areas, and formulating a power load pre-control scheme to perform electricity utilization prevention, control and treatment. The invention can solve the problems that the overload of the power grid is difficult to prevent, control and treat and the electricity consumption prediction is inaccurate in the winter of the returning country degree.

Inventors

PAN MIN
QI ZHENBIAO
ZHEN CHAO
ZHAO CHENG
GE JIAN
XU FEI
WANG YANG
ZHU BING
ZHANG YEMAO

Assignees

国网安徽省电力有限公司
国网安徽省电力有限公司电力科学研究院
安徽明生恒卓科技有限公司

Dates

Publication Date: 20260508
Application Date: 20250728

Claims (8)

1. A multi-factor correlated countryside winter power load pre-control method, the method comprising: s1, carrying out Fourier transformation on pre-acquired historical electricity utilization data to obtain power load data of a power grid, and carrying out load fluctuation identification on the power load data to obtain electricity utilization problems of the power grid; s2, performing pareto analysis on the cause condition of the electricity consumption problem to obtain influence factors of the electricity consumption problem, wherein the method comprises the following steps: Carrying out cause classification on the electricity utilization problem to obtain cause classification of the electricity utilization problem; carrying out quantitative statistics on the problems based on the cause categories to obtain the number of the problems of the cause categories; drawing a pareto chart of the electricity utilization problem based on the cause category and the problem number; performing correlation verification on the electricity utilization problem based on the pareto graph to obtain influence factors of the electricity utilization problem; Carrying out load characteristic cluster analysis on the influence factors to obtain a strong correlation factor data matrix of the electricity utilization problem, wherein the method comprises the following steps: inputting the influence factors into a load characteristic clustering model to obtain clustering data of the electricity utilization problem; and performing influence calculation based on the clustering data and a forward influence formula to obtain the forward influence of the influence factor, wherein the forward influence formula is as follows: in the formula, For the number of the historical electricity usage data, For the numbering of the cluster data, Is the first Historical electricity data The positive influence of the individual cluster data, Is the first Historical electricity data The content of the individual cluster data is selected, Is the first The minimum value of the individual cluster data, Is the first The maximum value of the individual cluster data is, Is the first Clustering data; and carrying out influence calculation based on the clustering data and a negative influence formula to obtain the negative influence of the influence factors, wherein the negative influence formula is as follows: in the formula, For the number of the historical electricity usage data, For the numbering of the cluster data, Is the first Historical electricity data The negative influence of the individual cluster data, Is the first Historical electricity data The content of the individual cluster data is selected, Is the first The minimum value of the individual cluster data, Is the first The maximum value of the individual cluster data is, Is the first Clustering data; establishing the strong correlation factor data matrix based on the positive influence and the negative influence; Carrying out quantization analysis on the strong correlation factor data matrix to obtain a correlation influence coefficient of the power grid; s3, carrying out iterative training on the power load prediction model based on the association influence coefficient to obtain a trained power load prediction model, and inputting current power consumption data obtained in advance into the trained power load prediction model to obtain load prediction data of the power grid; s4, classifying colors of the power grid based on the load condition of the load prediction data and the severity of the electricity utilization problem to obtain a red-yellow-green classification area of the power grid; And S5, formulating a power load pre-control scheme of the winter in the rural area based on the red-yellow-green grading area, and performing electricity utilization prevention and control treatment on the power grid according to the power load pre-control scheme.
2. The multi-factor correlated power load pre-control method for winter in countryside of claim 1, wherein the fourier transforming the pre-acquired historical power consumption data to obtain the power load data of the power grid comprises: performing noise cleaning on the historical electricity consumption data to obtain standardized data of the historical electricity consumption data; periodically dividing the standardized data to obtain segmented data of the standardized data; performing Fourier transform on the segmented data to obtain electric power data of the segmented data; and extracting load characteristics of the power data to obtain the power load data of the power grid.
3. The multi-factor correlated countryside winter power load pre-control method of claim 1, wherein the performing load fluctuation identification on the power load data to obtain the power consumption problem of the power grid comprises the following steps: Performing feature extraction on time domain data and frequency domain data of the power load data to obtain time-frequency features of the power load data; Carrying out abnormal fluctuation identification on the time-frequency characteristics to obtain abnormal fluctuation data of the time-frequency characteristics; carrying out statistical characteristic analysis on the historical electricity consumption data to obtain a load fluctuation threshold value of the power grid; and performing problem diagnosis on the power grid based on the abnormal fluctuation data and the load fluctuation threshold value to obtain the electricity utilization problem of the power grid.
4. The method for pre-controlling the power load of the rural power grid in the countryside of multi-factor association according to claim 1, wherein the performing quantization analysis on the strong association factor data matrix to obtain the association influence coefficient of the power grid comprises the following steps: Trend analysis is carried out on the strong correlation factor data matrix to obtain the change trend of the strong correlation factor data matrix; Performing dispersion analysis on the strong-correlation factor data matrix to obtain the dispersion degree of the strong-correlation factor data matrix; and carrying out association relation analysis on the strong association factor data matrix based on the change trend and the discrete degree to obtain an association influence coefficient of the power grid.
5. The method for pre-controlling the power load of the rural power grid in the winter with multi-factor correlation as set forth in claim 1, wherein the iterative training of the power load prediction model based on the correlation influence coefficient to obtain a trained power load prediction model, inputting the pre-acquired current power consumption data into the trained power load prediction model to obtain the load prediction data of the power grid, comprises: Performing iterative training on the power load prediction model based on the association influence coefficient to obtain the power load prediction model after iteration; and inputting the historical electricity consumption data into the iterative power load prediction model to calculate a loss value to obtain the loss value of the iterative power load prediction model, wherein the iterative loss value calculation function is as follows: in the formula, As a result of the value of the loss, For the amount of the historical electricity usage data, For the number of associated influence coefficients, For the number of the historical electricity usage data, For the number of the associated influence coefficient, Is the first The influence coefficients are associated with each other, Is based on the first The power load of the calculation of the respective associated influence coefficient, Is the first A history of the use of electricity data, Is the first True values of the historical electricity data; Determining a trained power load prediction model according to the loss value; and inputting the pre-acquired current electricity utilization data into the trained power load prediction model to obtain load prediction data of the power grid.
6. The multi-factor correlated reverse winter power load pre-control method as set forth in claim 1, wherein the classifying color division is performed on the power grid based on the load condition of the load prediction data and the severity of the power utilization problem to obtain a red-yellow-green classification area of the power grid, comprising: Load classification is carried out on the load prediction data, so that the predicted load classification of the power grid is obtained; the severity degree screening is carried out on the electricity utilization problems, so that the serious electricity utilization problems of the power grid are obtained; Establishing a quantitative relation between the predicted load classification and the serious electricity utilization problem; And classifying colors of the power grid according to the quantization relation to obtain a red-yellow-green classification area of the power grid.
7. The multi-factor correlated power load pre-control method for the rural power winter according to claim 6, wherein the power load pre-control scheme for the rural power winter based on the red-yellow-green classification area is formulated, and the power grid is subjected to power utilization prevention and control treatment according to the power load pre-control scheme, and the method comprises the following steps: monitoring the red-yellow-green classification platform area to obtain real-time load data of the red-yellow-green classification platform area; performing demand analysis on the red-yellow-green classification transformer area based on the real-time load data and the quantization relation to obtain the electricity consumption demand of the red-yellow-green classification transformer area; Formulating a power load pre-control scheme of the winter in the rural area based on the power demand and the pre-acquired distributed resource data; And carrying out electricity utilization prevention and control treatment on the power grid according to the electric load pre-control scheme.
8. A multi-factor correlated countryside winter power load pre-control system, the system comprising: The power consumption problem identification module is used for carrying out Fourier transformation on the pre-acquired historical power consumption data to obtain power load data of a power grid, and carrying out load fluctuation identification on the power load data to obtain the power consumption problem of the power grid; The power load influence factor analysis module is used for performing pareto analysis on the cause condition of the electricity consumption problem to obtain the influence factor of the electricity consumption problem, and comprises the following components: Carrying out cause classification on the electricity utilization problem to obtain cause classification of the electricity utilization problem; carrying out quantitative statistics on the problems based on the cause categories to obtain the number of the problems of the cause categories; drawing a pareto chart of the electricity utilization problem based on the cause category and the problem number; performing correlation verification on the electricity utilization problem based on the pareto graph to obtain influence factors of the electricity utilization problem; Carrying out load characteristic cluster analysis on the influence factors to obtain a strong correlation factor data matrix of the electricity utilization problem, wherein the method comprises the following steps: inputting the influence factors into a load characteristic clustering model to obtain clustering data of the electricity utilization problem; and performing influence calculation based on the clustering data and a forward influence formula to obtain the forward influence of the influence factor, wherein the forward influence formula is as follows: in the formula, For the number of the historical electricity usage data, For the numbering of the cluster data, Is the first Historical electricity data The positive influence of the individual cluster data, Is the first Historical electricity data The content of the individual cluster data is selected, Is the first The minimum value of the individual cluster data, Is the first The maximum value of the individual cluster data is, Is the first Clustering data; and carrying out influence calculation based on the clustering data and a negative influence formula to obtain the negative influence of the influence factors, wherein the negative influence formula is as follows: in the formula, For the number of the historical electricity usage data, For the numbering of the cluster data, Is the first Historical electricity data The negative influence of the individual cluster data, Is the first Historical electricity data The content of the individual cluster data is selected, Is the first The minimum value of the individual cluster data, Is the first The maximum value of the individual cluster data is, Is the first Clustering data; establishing the strong correlation factor data matrix based on the positive influence and the negative influence; Carrying out quantization analysis on the strong correlation factor data matrix to obtain a correlation influence coefficient of the power grid; The power load prediction module is used for carrying out iterative training on the power load prediction model based on the association influence coefficient to obtain a trained power load prediction model, and inputting current power utilization data obtained in advance into the trained power load prediction model to obtain load prediction data of the power grid; The grading platform region dividing module is used for carrying out grading color division on the power grid based on the load condition of the load prediction data and the severity of the electricity utilization problem to obtain a red-yellow-green grading platform region of the power grid; And the power load pre-control module is used for making a power load pre-control scheme of the rural power returning degree winter based on the red, yellow and green grading areas and carrying out electricity utilization prevention and control treatment on the power grid according to the power load pre-control scheme.

Description

Multi-factor-associated method and system for pre-controlling power load of winter in rural area Technical Field The invention relates to the technical field of smart grids, in particular to a multi-factor-associated method and system for pre-controlling power load of a rural power plant in winter. Background The power load pre-control system for the winter in the countryside degree mainly comprises a power load prevention and control scheme and a power load prevention and control prediction model. However, the traditional prevention and control scheme mainly depends on manual experience and simple threshold setting, lacks an intelligent decision support system, cannot timely and accurately judge the reason and trend of load increase when the load exceeds a preset threshold, is difficult to develop a strong-pertinence scientific and reasonable prevention and control strategy, and has weak cooperative prevention and control capability among power systems in different areas, and in addition, the information transmission is not timely in the transregional power allocation process, the coordination mechanism is imperfect, so that high-efficiency load prevention and control are difficult to realize, and the stability and safety of power supply in the winter of returning rural power cannot be ensured. The existing power load prediction model is difficult to fully mine potential relation between the multi-source factors and the power load, so that the prediction accuracy is difficult to meet actual requirements. The traditional prediction method is mainly based on historical data modeling, dynamic and uncertain factors such as population flow are not considered sufficiently, so that a large deviation exists between a prediction result and an actual load condition, meanwhile, meteorological factors have key influence on winter power load, but the diversity of meteorological conditions and the complexity of meteorological changes in different areas further increase the difficulty of accurate prediction. Disclosure of Invention The invention provides a multi-factor-associated method and system for pre-controlling power load of a rural power grid in winter, which mainly aim to solve the problems that the overload of the power grid is difficult to prevent, control and treat and the power consumption prediction is inaccurate in the rural power grid in winter. In order to achieve the above purpose, the invention provides a multi-factor-associated method for pre-controlling power load of winter in rural area, which comprises the following steps: s1, carrying out Fourier transformation on pre-acquired historical electricity utilization data to obtain power load data of a power grid, and carrying out load fluctuation identification on the power load data to obtain electricity utilization problems of the power grid; S2, performing pareto analysis on the cause condition of the electricity consumption problem to obtain an influence factor of the electricity consumption problem, performing load characteristic cluster analysis on the influence factor to obtain a strong association factor data matrix of the electricity consumption problem, and performing quantitative analysis on the strong association factor data matrix to obtain an association influence coefficient of the power grid; s3, carrying out iterative training on the power load prediction model based on the association influence coefficient to obtain a trained power load prediction model, and inputting current power consumption data obtained in advance into the trained power load prediction model to obtain load prediction data of the power grid; s4, classifying colors of the power grid based on the load condition of the load prediction data and the severity of the electricity utilization problem to obtain a red-yellow-green classification area of the power grid; And S5, formulating a power load pre-control scheme of the winter in the rural area based on the red-yellow-green grading area, and performing electricity utilization prevention and control treatment on the power grid according to the power load pre-control scheme. In a preferred embodiment, the fourier transforming the pre-acquired historical electricity data to obtain the power load data of the power grid includes: performing noise cleaning on the historical electricity consumption data to obtain standardized data of the historical electricity consumption data; periodically dividing the standardized data to obtain segmented data of the standardized data; performing Fourier transform on the segmented data to obtain electric power data of the segmented data; and extracting load characteristics of the power data to obtain the power load data of the power grid. In a preferred embodiment, the load fluctuation identification of the power load data, to obtain the electricity consumption problem of the power grid, includes: Performing feature extraction on time domain data and frequency domain data of the power load d