A power grid intelligent dynamic power-saving dispatching method and system

Abstract

The application belongs to the technical field of power saving scheduling, and provides a power grid intelligent dynamic power saving scheduling method and system, which comprises the following steps: analyzing the spatial correlation degree of an abnormal power consumption area and adjacent abnormal power consumption areas, evaluating the abnormal power consumption correlation degree of the abnormal power consumption area and the adjacent abnormal power consumption areas in the geographical space, and determining a high correlation area group; analyzing the power consumption complementarity between the abnormal power consumption areas in the high correlation area group, and judging the power consumption complementarity between the abnormal power consumption areas in the high correlation area group; analyzing the time dislocation between the abnormal power consumption areas in the complementary area group, determining the complementary time period characteristics between the abnormal power consumption areas in the complementary area group, and constructing a complementary time period prediction model; through complementarity analysis and prediction modeling, the application realizes fine power coordination optimization of the high correlation abnormal power consumption area, effectively balances the regional power consumption peak and valley, reduces power abandonment, and improves the power grid load rate.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of power-saving scheduling, in particular to a power grid intelligent dynamic power-saving scheduling method and system. BACKGROUND

[0002] With the rapid growth of power demand and large-scale access of renewable energy, the power grid operation is facing problems such as high peak load pressure, resource waste in low valley period, and difficulty in accurately regulating abnormal power consumption behavior.

[0003] In the prior art, the complementary analysis lacks a space-time dynamic matching mechanism, which makes it difficult to accurately connect the peak and low valley periods; the prediction model mostly uses traditional statistical methods or fixed rule libraries, which cannot capture the nonlinear characteristics of time series; the control strategy of energy storage devices is static and does not link with the prediction period, resulting in low resource utilization. The above problems lead to insufficient grid scheduling accuracy and response lag, which is difficult to meet the dynamic power-saving demand.

[0004] Therefore, the application provides a power grid intelligent dynamic power-saving scheduling method and system. SUMMARY

[0005] In order to make up for the deficiencies of the prior art and solve at least one technical problem proposed in the background art.

[0006] In a first aspect, the application provides a power grid intelligent dynamic power-saving scheduling method, comprising:

[0007] identifying abnormal power consumption areas in the power supply area through user power consumption data in the power supply area of the power grid;

[0008] By analyzing the spatial correlation degree of the abnormal power consumption area and the adjacent abnormal power consumption area, the abnormal power consumption correlation degree of the abnormal power consumption area and the adjacent abnormal power consumption area in the geographical space is evaluated, and a high correlation area group is determined;

[0009] By analyzing the power consumption complementarity between the abnormal power consumption areas in the high correlation area group, the power consumption complementarity between the abnormal power consumption areas in the high correlation area group is determined;

[0010] If the power consumption complementarity between the abnormal power consumption areas in the high correlation area group is high, it is recorded as a complementary area group, and by analyzing the time dislocation of the abnormal power consumption areas in the complementary area group, the complementary time period characteristics between the abnormal power consumption areas in the complementary area group are determined, and a complementary time period prediction model is constructed. The complementary time period characteristics of the complementary area group are used as the input of the complementary time period prediction model, and the predicted complementary overlapping time period of the complementary area group is output.

[0011] Preferably, the specific process of identifying abnormal power consumption areas in the power supply area is as follows:

[0012] The user power consumption data includes user power consumption, and a power supply area of a power supply network is divided into a plurality of power consumption areas according to transformer areas; if the user power consumption in a power consumption area is not within a user power consumption range, a corresponding user is recorded as an abnormal user;

[0013] The number of abnormal users in the power consumption area is calculated, and if the number of abnormal users is greater than an abnormal user number threshold, the corresponding power consumption area is recorded as an abnormal power consumption area.

[0014] Preferably, the specific process of evaluating the correlation degree of the abnormal power consumption area and the adjacent abnormal power consumption area in the geographical space is as follows:

[0015] Geographical boundary data of the abnormal power consumption area is obtained, the center coordinates of each adjacent abnormal power consumption area of the abnormal power consumption area are calculated, coordinate system conversion is performed, the straight-line distance between the center point of the abnormal power consumption area and the adjacent abnormal power consumption area is calculated, and a Gaussian function is used to model the spatial correlation degree between the i-th abnormal power consumption area and the j-th abnormal power consumption area;

[0016] If the spatial correlation degree is greater than or equal to a spatial correlation degree threshold, it indicates that the spatial correlation degree between the abnormal power consumption area and the corresponding adjacent abnormal power consumption area is high, otherwise, it indicates that the spatial correlation degree is low.

[0017] Preferably, the specific process of determining the high-correlation area group is as follows:

[0018] Based on the high spatial correlation degree between the abnormal power consumption area and the corresponding adjacent abnormal power consumption area, the spatial correlation degrees between the abnormal power consumption area and all adjacent abnormal power consumption areas are sorted from large to small to obtain a spatial correlation degree sorting table of the abnormal power consumption area, the adjacent abnormal power consumption area located in the first order in the spatial correlation degree sorting table of the abnormal power consumption area is extracted, and the abnormal power consumption area and the adjacent abnormal power consumption area are integrated into a high-correlation area group.

[0019] Preferably, the specific process of judging the complementarity between the abnormal power consumption areas in the high-correlation area group is as follows:

[0020] In a monitoring period, the monitoring period is divided into a plurality of time intervals, the user power consumptions of the abnormal power consumption areas in the high-correlation area group in the monitoring period are summed to obtain the area power consumption of the monitoring period;

[0021] If the area power consumption of the monitoring period is higher than an area power consumption maximum value, the corresponding monitoring period is recorded as an abnormal peak period, and the area power consumption of the monitoring period is subtracted from the area power consumption maximum value to obtain a peak abnormal value; otherwise, the corresponding monitoring period is recorded as an abnormal trough period, and the area power consumption of the monitoring period is subtracted from an area power consumption minimum value, and the absolute value is taken to obtain a trough abnormal value;

[0022] The abnormal peak period, the peak abnormal value, the abnormal low peak period, and the low peak abnormal value are analyzed to determine the peak compensable period and the low peak compensable period, the total number of the peak compensable period and the low peak compensable period of the abnormal power consumption region in the high correlation region group in the monitoring period is counted, and ratio processing is performed on the total number of the abnormal peak period and the abnormal low peak period of the abnormal power consumption region in the high correlation region group to obtain a complementarity index. If the complementarity index is greater than a complementarity index threshold value, it indicates that the complementarity between the abnormal power consumption regions in the high correlation region group is high, otherwise, it indicates that the complementarity is low.

[0023] Preferably, the determination process of the peak compensable period is as follows:

[0024] The abnormal power consumption regions in the high correlation region group are respectively denoted as a first abnormal power consumption region and a second abnormal power consumption region, and the abnormal peak periods of the first abnormal power consumption region in the high correlation region group are integrated into an abnormal peak period sequence in chronological order;

[0025] The abnormal low peak periods of the second abnormal power consumption region in the high correlation region group are integrated into an abnormal low peak period sequence in chronological order;

[0026] Any abnormal peak period in the abnormal peak period sequence of the first abnormal power consumption region is extracted and combined with all abnormal low peak periods in the abnormal low peak period sequence of the second abnormal power consumption region one by one to construct a first period analysis group;

[0027] The peak abnormal value corresponding to the abnormal peak period in the first period analysis group is subtracted from the low peak abnormal value corresponding to the abnormal low peak period, and the absolute value is taken to obtain a first peak-valley abnormal deviation value;

[0028] If the first peak-valley abnormal deviation value is less than or equal to a peak-valley abnormal deviation standard value, the abnormal peak period in the first period analysis group is recorded as a peak compensable period.

[0029] Preferably, the determination process of the low peak compensable period is as follows:

[0030] The abnormal low peak periods of the first abnormal power consumption region in the high correlation region group are integrated into an abnormal low peak period sequence in chronological order;

[0031] The abnormal peak periods of the second abnormal power consumption region in the high correlation region group are integrated into an abnormal peak period sequence in chronological order;

[0032] Any abnormal low peak period in the abnormal low peak period sequence of the first abnormal power consumption region is extracted and combined with all abnormal peak periods in the abnormal peak period sequence of the second abnormal power consumption region one by one to construct a second period analysis group;

[0033] Subtracting the low-peak abnormal value corresponding to the abnormal low-peak period in the second period analysis group from the high-peak abnormal value corresponding to the abnormal high-peak period, and taking the absolute value, a second peak-valley abnormal deviation value is obtained.

[0034] If the second peak-valley abnormal deviation value is less than or equal to the peak-valley abnormal deviation standard value, the abnormal low-peak period in the second period analysis group is recorded as a low-peak compensable period.

[0035] Preferably, the specific process of determining the complementary period characteristics between the abnormal power consumption regions in the compensable region group is as follows:

[0036] For any abnormal power consumption region (A, B) in the compensable region group, the region power consumption of the abnormal power consumption region A and the abnormal power consumption region B in the monitoring period is extracted, and is integrated into a region power consumption sequence A and a region power consumption sequence B in time sequence, respectively.

[0037] For the region power consumption sequence, the autocorrelation coefficient max of lag k=1, 2, …, k is calculated, and k max is set as the length of the monitoring period.

[0038] A curve graph of the autocorrelation coefficient is drawn, and whether a peak value appears at a certain lag and the peak value exceeds a preset confidence interval is observed, and the corresponding k is taken as the best displacement amount of the abnormal power consumption region (A, B) in the compensable region group.

[0039] Displacement is performed according to the best displacement amount, and a complementary overlap period and a power consumption difference value are determined.

[0040] The best displacement amount, the complementary overlap period and the power consumption difference value of the abnormal power consumption region A and the abnormal power consumption region B in the compensable region group are taken as the complementary period characteristics.

[0041] Preferably, the process of determining the complementary overlap period and the power consumption difference value is as follows:

[0042] The overlapping period of the abnormal power consumption region A and the abnormal power consumption region B in the compensable region group is counted, and is recorded as the complementary overlap period of the compensable region group.

[0043] The high-peak abnormal value and the low-peak abnormal value of the complementary overlap period in the compensable region group are subtracted, and the absolute value is taken, so that a power consumption difference value is obtained.

[0044] In a second aspect, the present application also provides a power grid intelligent dynamic power saving scheduling system, which comprises:

[0045] A data acquisition and analysis module: through the user power consumption data in the power supply region of the power grid, an abnormal power consumption region in the power supply region is identified.

[0046] Region correlation analysis module: through the analysis of the spatial correlation degree of the abnormal power consumption region and the adjacent abnormal power consumption region, the abnormal power consumption correlation degree of the abnormal power consumption region and the adjacent abnormal power consumption region in the geographical space is evaluated, and a high correlation region group is determined;

[0047] Complementarity analysis module: through the analysis of the power consumption complementarity between the abnormal power consumption regions in the high correlation region group, the power consumption complementarity between the abnormal power consumption regions in the high correlation region group is judged;

[0048] Complementary overlap period prediction module: if the power consumption complementarity between the abnormal power consumption regions in the high correlation region group is high, it is recorded as a complementary region group, through the analysis of the time dislocation of the abnormal power consumption regions in the complementary region group, the complementary time period characteristics between the abnormal power consumption regions in the complementary region group are determined, and a complementary time period prediction model is constructed, the complementary time period characteristics of the complementary region group are taken as the input of the complementary time period prediction model, and the predicted complementary overlap period of the complementary region group is output.

[0049] The beneficial effects of the present application are as follows:

[0050] 1、The present application improves the positioning accuracy through the objective quantitative identification of the abnormal power consumption region; the spatial correlation degree between the abnormal regions is scientifically quantified by combining GIS geographic information and Gaussian function model, the correlation strength is dynamically adjusted through the distance attenuation coefficient, and the high correlation region group is extracted based on the sorting strategy, which provides a structured basis for subsequent complementarity analysis and resource coordination, and provides key technical support for optimizing power grid dispatching and reducing operation risk.

[0051] 2、The present application realizes fine power coordination optimization of high correlation abnormal power consumption region through complementarity analysis and prediction modeling, determines the best dislocation amount based on time dislocation analysis, constructs a prediction model combining LSTM model, can predict the complementary overlap period in advance, improves the predictability of the dispatching strategy; through the charge and discharge control of the energy storage device in the prediction period, the regional power consumption peak and valley are effectively balanced, the power abandonment is reduced and the power grid load rate is improved, which is beneficial to improve the stability and operation efficiency of the power system. BRIEF DESCRIPTION OF DRAWINGS

[0052] The present application will be further described below in conjunction with the drawings.

[0053] Figure 1 It is a step flow chart of a power grid intelligent dynamic power saving dispatching method according to an embodiment of the present application;

[0054] Figure 2 It is a system block diagram of a power grid intelligent dynamic power saving dispatching system according to an embodiment of the present application. DETAILED DESCRIPTION

[0055] In order to make the technical means, creative features, purposes and effects of the present application easy to understand, the present application will be further described below in conjunction with specific embodiments. Embodiment 1

[0056] Please refer to Figure 1 The power grid intelligent dynamic power-saving scheduling method of the embodiment of the present application includes the following steps:

[0057] Step one: identify the abnormal power consumption area in the power supply area through the user power consumption data in the power supply area of the power grid;

[0058] The user power consumption data includes the user power consumption;

[0059] The power supply area of the power grid is divided into several power consumption areas according to the transformer area. The transformer area is a power supply unit in the power system with the distribution transformer as the core. The accurate range of the target transformer area needs to be determined through the GIS (Geographic Information System) of the power enterprise, including the user list covered, transformer parameters, line topology, etc.

[0060] Through the concentrator and the smart meter, the user power consumption data in the power consumption area is collected in real time, and the user power consumption data is preprocessed, including using Z-Score standardization for normalization processing to eliminate the dimension influence, and filtering the abnormal values through the box plot to eliminate the obvious abnormal data.

[0061] Based on the user power consumption of the power consumption area, the user power consumption is compared with the user power consumption range:

[0062] If the user power consumption in the power consumption area is not within the user power consumption range, the corresponding user is recorded as an abnormal user;

[0063] If the user power consumption in the power consumption area is within the user power consumption range, the corresponding user is recorded as a normal user;

[0064] It should be noted that the user power consumption range is set by the workers in the field according to the characteristics of the user power consumption data combined with historical experience;

[0065] The number of abnormal users in the power consumption area is counted, and the proportion of the number of abnormal users in the power consumption area is calculated;

[0066] In some embodiments, the proportion of the number of abnormal users is compared with the proportion threshold of the number of abnormal users:

[0067] If the proportion of the number of abnormal users is greater than the proportion threshold of the number of abnormal users, the corresponding power consumption area is recorded as an abnormal power consumption area;

[0068] If the proportion of the number of abnormal users is less than or equal to the threshold of the proportion of the number of abnormal users, the corresponding power consumption area is recorded as a normal power consumption area;

[0069] It can be understood that the abnormal power consumption area refers to that the user power consumption in the power consumption area is out of the user power consumption range or is lower than the user power consumption range, and a power consumption peak or a power consumption trough occurs;

[0070] Step two: by analyzing the spatial correlation degree of the abnormal power consumption area and the adjacent abnormal power consumption area, the abnormal power consumption correlation degree of the abnormal power consumption area and the adjacent abnormal power consumption area in the geographical space is evaluated, and a high correlation area group is determined;

[0071] It should be noted that the adjacent abnormal power consumption area refers to the area close to the abnormal power consumption area in the geographical space;

[0072] Based on any abnormal power consumption area, the adjacent abnormal power consumption area of the abnormal power consumption area is extracted, and the abnormal power consumption correlation degree of the abnormal power consumption area and the adjacent abnormal power consumption area in space is analyzed, specifically:

[0073] A1, obtaining geographical boundary data of the abnormal power consumption area: based on the GIS geographic information system of the power enterprise, the geographical boundary polygon vertex coordinates of two abnormal power consumption areas are extracted, which are usually stored in the form of latitude and longitude (degree, minute and second), representing the spatial range of the area (for example, a polygon boundary formed by multiple vertices of a transformer area) ;

[0074] A2, calculating the center coordinates of each adjacent abnormal power consumption area of the abnormal power consumption area: in the GIS geographic information system, for the abnormal power consumption area with regular shape, the center of the abnormal power consumption area is determined by the geometric center calculation method; for the area with irregular shape, the center point is calculated by using the barycentric algorithm, that is, the position of the center point is determined by calculating the weighted average value of the coordinates of all points in the area;

[0075] A3, coordinate system conversion: since the original coordinates in GIS are usually latitude and longitude, there will be errors in direct distance calculation, so it is necessary to convert latitude and longitude into plane rectangular coordinates to ensure the accuracy of distance calculation;

[0076] A4, calculating the straight line distance between the center point of the abnormal power consumption area and the adjacent abnormal power consumption area: using the distance measurement tool in the GIS software, the straight line distance between the center points of the two abnormal power consumption areas is directly measured;

[0077] The closer the spatial distance, the stronger the potential correlation, and according to the straight line distance between the center point of the abnormal power consumption area and the adjacent abnormal power consumption area, the spatial correlation degree between the i-th abnormal power consumption area and the j-th abnormal power consumption area is modeled by using the Gaussian function :

[0078]

[0079] In the formula, d is the straight-line distance between the center points of the i-th abnormal power consumption area and the j-th abnormal power consumption area, is a distance attenuation coefficient, which is set according to the scale of the power supply network;

[0080] In some embodiments, the spatial correlation degree is compared with a spatial correlation degree threshold value:

[0081] If the spatial correlation degree is greater than or equal to the spatial correlation degree threshold value, it indicates that the spatial correlation degree between the abnormal power consumption area and the corresponding adjacent abnormal power consumption area is high;

[0082] If the spatial correlation degree is less than the spatial correlation degree threshold value, it indicates that the spatial correlation degree between the abnormal power consumption area and the corresponding adjacent abnormal power consumption area is low;

[0083] Based on the high spatial correlation degree between the abnormal power consumption area and the corresponding adjacent abnormal power consumption area, the spatial correlation degrees between the abnormal power consumption area and all adjacent abnormal power consumption areas are sorted from large to small to obtain a spatial correlation degree sorting table of the abnormal power consumption area, and the adjacent abnormal power consumption area located in the first order in the spatial correlation degree sorting table of the abnormal power consumption area is extracted and integrated with the abnormal power consumption area as a high correlation area group;

[0084] It can be understood that the high and low spatial correlation degrees between the abnormal power consumption area and the corresponding adjacent abnormal power consumption area facilitate power consumption scheduling in geographical space, provide reliable support for subsequent complementary analysis of the abnormal power consumption area, and are further conducive to scheduling abnormal power consumption according to the prediction result;

[0085] The technical scheme of the embodiment is: identifying an abnormal power consumption area in a power supply area in a power supply network through user power consumption data in the power supply area, analyzing the spatial correlation degrees of the abnormal power consumption area and adjacent abnormal power consumption areas, evaluating the abnormal power consumption correlation degrees of the abnormal power consumption area and the adjacent abnormal power consumption areas in geographical space, and determining a high correlation area group. The application improves the positioning accuracy through objective quantitative identification of the abnormal power consumption area. The spatial correlation degrees between abnormal areas are scientifically quantified by combining GIS geographical information and a Gaussian function model, the correlation strength is dynamically adjusted through a distance attenuation coefficient, and a high correlation area group is extracted based on a sorting strategy, which provides a structured basis for subsequent complementary analysis and resource coordination, and provides key technical support for optimizing power grid scheduling and reducing operation risks. Embodiment 2

[0086] Please refer to Figure 1 The power supply network intelligent dynamic power saving scheduling method of the embodiment of the application further includes the following steps:

[0087] Step three: judging the electricity complementarity between the abnormal electricity regions in the high correlation region group by analyzing the electricity complementarity between the abnormal electricity regions in the high correlation region group;

[0088] In the monitoring period, the monitoring period is divided into several time intervals, and the electricity consumption of all users in the abnormal electricity region in the high correlation region group in the monitoring period is summed to obtain the regional electricity consumption of the monitoring period;

[0089] If the regional electricity consumption of the monitoring period is higher than the maximum regional electricity consumption, the corresponding monitoring period is recorded as an abnormal peak period, and the regional electricity consumption of the monitoring period is subtracted from the maximum regional electricity consumption to obtain a peak abnormal value;

[0090] If the regional electricity consumption of the monitoring period is lower than the minimum regional electricity consumption, the corresponding monitoring period is recorded as an abnormal low peak period, and the regional electricity consumption of the monitoring period is subtracted from the minimum regional electricity consumption, and the absolute value is taken to obtain a low peak abnormal value;

[0091] It should be noted that the regional electricity consumption extreme value includes the maximum regional electricity consumption and the minimum regional electricity consumption, and both are set by the person skilled in the art according to historical experience;

[0092] The abnormal electricity regions in the high correlation region group are recorded as the first abnormal electricity region and the second abnormal electricity region respectively;

[0093] The abnormal peak period of the first abnormal electricity region in the high correlation region group is extracted, and the complementarity analysis is performed with the abnormal low peak period of the second abnormal electricity region in the high correlation region group, specifically:

[0094] The abnormal peak period of the first abnormal electricity region in the high correlation region group is integrated into an abnormal peak period sequence in time order , n represents the total number of abnormal peak periods in the first abnormal electricity region;

[0095] The abnormal low peak period of the second abnormal electricity region in the high correlation region group is integrated into an abnormal low peak period sequence in time order , m represents the total number of abnormal low peak periods in the second abnormal electricity region;

[0096] Any abnormal peak period in the abnormal peak period sequence of the first abnormal electricity region is extracted, and all abnormal low peak periods in the abnormal low peak period sequence of the second abnormal electricity region are combined one by one to construct a first time period analysis group;

[0097] The peak abnormal value corresponding to the abnormal peak period in the first time period analysis group is subtracted from the low peak abnormal value corresponding to the abnormal low peak period, and the absolute value is taken to obtain a first peak-valley abnormal deviation value;

[0098] The first peak-valley abnormal deviation value is compared with the peak-valley abnormal deviation standard value:

[0099] If the first peak-valley abnormal deviation value is greater than the peak-valley abnormal deviation standard value, the combination is cancelled;

[0100] If the first peak-valley abnormal deviation value is less than or equal to the peak-valley abnormal deviation standard value, the abnormal peak period in the first period analysis group is recorded as a peak-supplementable period;

[0101] It should be noted that the peak-valley abnormal deviation standard value is set by a person skilled in the art according to historical experience;

[0102] Similarly, the abnormal off-peak period of the first abnormal power consumption area in the high-correlation area group is extracted, and the complementary analysis is performed with the abnormal peak period of the second abnormal power consumption area in the high-correlation area group, in particular:

[0103] The abnormal off-peak period of the first abnormal power consumption area in the high-correlation area group is integrated into an abnormal off-peak period sequence in chronological order , r represents the total number of abnormal off-peak periods in the abnormal power consumption area;

[0104] The abnormal peak period of the second abnormal power consumption area in the high-correlation area group is integrated into an abnormal peak period sequence in chronological order , k represents the total number of abnormal peak periods in the abnormal power consumption area;

[0105] Any abnormal off-peak period in the abnormal off-peak period sequence of the first abnormal power consumption area is extracted, and one-to-one combination is performed with all abnormal peak periods in the abnormal peak period sequence of the second abnormal power consumption area, to construct a second period analysis group;

[0106] The low-peak abnormal value corresponding to the abnormal off-peak period in the second period analysis group is subtracted from the high-peak abnormal value corresponding to the abnormal peak period, and the absolute value is taken, to obtain a second peak-valley abnormal deviation value;

[0107] The second peak-valley abnormal deviation value is compared with the peak-valley abnormal deviation standard value:

[0108] If the second peak-valley abnormal deviation value is greater than the peak-valley abnormal deviation standard value, the combination is cancelled;

[0109] If the second peak-valley abnormal deviation value is less than or equal to the peak-valley abnormal deviation standard value, the abnormal off-peak period in the second period analysis group is recorded as a low-peak-supplementable period;

[0110] The total number of peak-supplementable periods and low-peak-supplementable periods of the abnormal power consumption area in the high-correlation area group in the monitoring period is counted, and ratio processing is performed with the total number of abnormal peak periods and abnormal off-peak periods of the abnormal power consumption area in the high-correlation area group, to obtain a complementarity index;

[0111] In some embodiments, the complementarity index is compared with a complementarity index threshold:

[0112] If the complementarity index is greater than the complementarity index threshold, it indicates that the complementarity between abnormal power consumption areas in the highly correlated area group is high.

[0113] If the complementarity index is less than or equal to the complementarity index threshold, it indicates that the complementarity between abnormal power consumption areas in the highly correlated area group is low.

[0114] It is understandable that the complementarity index threshold is set by those skilled in the art based on expert experience and industry standards, and can be set to 0.75;

[0115] The purpose of setting a complementarity index is:

[0116] Function 1: The complementarity index quantifies the degree of complementarity between two abnormal power consumption areas in the time dimension by using the ratio of the number of compensable time periods to the total number of abnormal time periods. The higher the index value, the more the abnormal peak and off-peak periods overlap between the two areas, the smaller the abnormal deviation, and the stronger the complementarity.

[0117] Function 2: The complementarity index reflects the ability of highly correlated regional groups to cover abnormal electricity consumption periods. A higher complementarity index means that most abnormal periods can be mitigated through complementarity, requiring only additional measures (such as demand response and energy storage release) for a few periods.

[0118] Step 4: If the electricity complementarity between abnormal electricity consumption areas in the highly correlated area group is high, it is recorded as a complementary area group. By analyzing the time misalignment of abnormal electricity consumption areas in the complementary area group, the complementary time period characteristics between abnormal electricity consumption areas in the complementary area group are determined, and a complementary time period prediction model is constructed. The complementary time period characteristics of the complementary area group are used as the input of the complementary time period prediction model, and the predicted complementary overlapping time periods of the complementary area group are output.

[0119] For any abnormal power consumption area (A, B) in a complementary area group, extract the regional power consumption of abnormal power consumption area A and abnormal power consumption area B in the complementary area group during the monitoring period, and integrate them into regional power consumption sequence A and regional power consumption sequence B respectively in chronological order;

[0120] For regional electricity consumption sequences, the following is used: Calculate the lag k=1,2,…,k max autocorrelation coefficient k max The duration corresponding to the monitoring cycle is set in the formula. This represents the observed value at time point t in the regional electricity consumption sequence A. A Bt represents the observation value of the tth time point in the region electricity consumption sequence B, A represents the mean of all observation values in the region electricity consumption sequence B, n represents the total number of time points within the monitoring period;

[0121] Draw the autocorrelation coefficient curve, observe whether there is a peak value at a certain lag, and the peak value exceeds the preset confidence interval, then take the corresponding k as the best misalignment amount of the abnormal electricity consumption region (A, B) in the complementary region group;

[0122] It should be noted that the purpose of determining the best misalignment amount of the abnormal electricity consumption region (A, B) in the complementary region group is to quantize and locate the moment of the strongest reverse correlation of the electricity consumption patterns of the two abnormal electricity consumption regions through mathematical methods, thereby providing a key basis for subsequent complementary period prediction and power resource optimization. The best misalignment amount k directly corresponds to the time offset of the complementary period, and the best misalignment amount k is one of the core features of the complementary period prediction model; for example: in the historical data, if the complementary period of region A and B usually occurs at k =−2 (2 hours ahead), the prediction model can learn this rule and infer the time offset of the future complementary period. Combined with other features (such as electricity consumption difference), the model can output the specific time range of the future complementary period (such as 2023-08-01 6:00-8:00). By predicting the complementary period, the power company can adjust the power supply strategy in advance, for example: delivering peak power of the abnormal electricity consumption region A to the valley period of the abnormal electricity consumption region B, reducing the demand for abandoned electricity or energy storage;

[0123] After misaligning according to the best misalignment amount, the overlapping period of the peak complementary period and the low peak complementary period of the abnormal electricity consumption region A and the abnormal electricity consumption region B in the complementary region group is counted, which is recorded as the complementary overlapping period of the complementary region group;

[0124] After misaligning according to the best misalignment amount, the peak abnormal value and the low peak abnormal value of the complementary overlapping period in the complementary region group are processed by difference, and the absolute value is taken, to obtain the electricity consumption difference value;

[0125] The best misalignment amount, the complementary overlapping period, and the electricity consumption difference value of the abnormal electricity consumption region A and the abnormal electricity consumption region B in the complementary region group are taken as the complementary period features, and a complementary period prediction model is constructed based on the LSTM (Long Short Term Memory Network) model. The complementary period features of the best misalignment amount, the complementary overlapping period, and the electricity consumption difference value of the two abnormal electricity consumption regions in the complementary region group in the historical period are divided into a training set (70%), a validation set (20%), and a test set (10%) in chronological order;

[0126] Feature scaling: standardization (Z-score) is performed on continuous features (such as misalignment amount, electricity consumption difference);

[0127] Model training: fit the model using the training set, adjust the hyperparameters (the number of hidden layer units of LSTM) through the validation set;

[0128] Error evaluation: use the test set to calculate the mean square error (MSE), mean absolute error (MAE) and other indicators to evaluate the performance of the model;

[0129] Take the complementary time period characteristics of the complementary region group in the current monitoring period as input, and output the predicted complementary overlap period;

[0130] In the predicted complementary overlap period, the two abnormal electricity consumption regions in the complementary region group are arranged according to the best staggering amount, the energy storage equipment is discharged in the peak complementary period to supplement the power supply gap, and the energy storage equipment is charged in the low peak complementary period to improve the grid load rate;

[0131] The technical scheme of the embodiment is: by analyzing the electricity complementarity between abnormal electricity consumption regions in the high correlation region group, determining the electricity complementarity between abnormal electricity consumption regions in the high correlation region group, if the electricity complementarity between abnormal electricity consumption regions in the high correlation region group is high, it is recorded as a complementary region group, by analyzing the time staggering of abnormal electricity consumption regions in the complementary region group, determining the complementary time period characteristics between abnormal electricity consumption regions in the complementary region group, and constructing a complementary time period prediction model, taking the complementary time period characteristics of the complementary region group as the input of the complementary time period prediction model, and outputting the predicted complementary overlap period of the complementary region group, the present application realizes the fine power coordination optimization of high correlation abnormal electricity consumption region through complementarity analysis and prediction modeling, determines the best staggering amount based on time staggering analysis, and constructs a prediction model combined with an LSTM model, which can predict the complementary overlap period in advance, improve the predictability of the scheduling strategy; through the charge and discharge control of the energy storage equipment in the predicted period, effectively balance the electricity consumption peak and valley of the region, reduce the abandoned electricity and improve the grid load rate, which is conducive to improving the stability and operation efficiency of the power system. Embodiment 3

[0132] Please refer to Figure 2 The power grid intelligent dynamic power saving scheduling system of the embodiment of the present application comprises the following modules:

[0133] Data acquisition and analysis module: identify the abnormal electricity consumption region in the power supply region through the user electricity consumption data in the power supply region;

[0134] Region correlation analysis module: analyze the spatial correlation of abnormal electricity consumption regions and adjacent abnormal electricity consumption regions, evaluate the abnormal electricity consumption correlation of abnormal electricity consumption regions and adjacent abnormal electricity consumption regions in geographical space, and determine the high correlation region group;

[0135] complementarity analysis module: by analyzing the electricity complementarity between the abnormal electricity consumption areas in the high correlation area group, the electricity complementarity between the abnormal electricity consumption areas in the high correlation area group is determined;

[0136] complementary overlap period prediction module: if the electricity complementarity between the abnormal electricity consumption areas in the high correlation area group is high, it is recorded as a complementary area group, by analyzing the time dislocation of the abnormal electricity consumption areas in the complementary area group, the complementary time period characteristics between the abnormal electricity consumption areas in the complementary area group are determined, and a complementary time period prediction model is constructed, the complementary time period characteristics of the complementary area group are taken as the input of the complementary time period prediction model, and the predicted complementary overlap period of the complementary area group is output.

[0137] The basic principles, main features and advantages of the present application are shown and described above. It should be understood by those skilled in the art that the present application is not limited by the above examples, and the above examples and descriptions in the specification are only to illustrate the principles of the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements all fall within the scope of the present application. The scope of protection of the present application is defined by the appended claims and their equivalents.

Claims

1. A power grid intelligent dynamic power saving dispatching method, characterized in that: The application comprises the following steps: Identify abnormal power consumption areas in the power supply area through user power consumption data in the power supply area; Analyze the spatial correlation degree of abnormal power consumption areas and adjacent abnormal power consumption areas to evaluate the abnormal power consumption correlation degree of abnormal power consumption areas and adjacent abnormal power consumption areas in geographical space and determine a high correlation area group; Analyze the power consumption complementarity between abnormal power consumption areas in the high correlation area group to determine the power consumption complementarity between abnormal power consumption areas in the high correlation area group; If the power consumption complementarity between abnormal power consumption areas in the high correlation area group is high, the high correlation area group is recorded as a complementary area group, the complementary time period characteristics of abnormal power consumption areas in the complementary area group are determined through time dislocation analysis of the complementary area group, and a complementary time period prediction model is constructed, the complementary time period characteristics of the complementary area group are taken as the input of the complementary time period prediction model, and the predicted complementary overlapping time period of the complementary area group is output.

2. The intelligent dynamic power-saving dispatching method for power grid according to claim 1, characterized in that: The specific process of identifying abnormal power consumption areas in the power supply area is as follows: User power consumption data includes user power consumption, and the power supply area of the power supply network is divided into several power consumption areas according to the transformer area; if the user power consumption in the power consumption area is not within the user power consumption range, the corresponding user is recorded as an abnormal user; Calculate the proportion of the number of abnormal users in the power consumption area; if it is greater than the abnormal user number proportion threshold, the corresponding power consumption area is recorded as an abnormal power consumption area.

3. The intelligent dynamic power-saving dispatching method for power grid according to claim 2, characterized in that: The specific process of evaluating the correlation degree of abnormal power consumption in geographical space between abnormal power consumption areas and adjacent abnormal power consumption areas is as follows: Obtain the geographical boundary data of the abnormal power consumption area, calculate the center coordinates of each adjacent abnormal power consumption area of the abnormal power consumption area, perform coordinate system conversion, calculate the straight line distance between the center point of the abnormal power consumption area and the adjacent abnormal power consumption area, and use the Gaussian function to model to obtain the spatial correlation degree; If greater than or equal to the spatial correlation degree threshold, it indicates that the spatial correlation degree between the abnormal power consumption area and the corresponding adjacent abnormal power consumption area is high, otherwise, the spatial correlation degree is low.

4. The intelligent dynamic power-saving dispatching method for power grid according to claim 3, characterized in that: The determination process of the high correlation area group is as follows: Based on the high spatial correlation degree between the abnormal power consumption area and the corresponding adjacent abnormal power consumption area, the spatial correlation degrees between the abnormal power consumption area and all adjacent abnormal power consumption areas are sorted from large to small to obtain a spatial correlation degree sorting table of the abnormal power consumption area, the adjacent abnormal power consumption area located in the first order in the spatial correlation degree sorting table of the abnormal power consumption area is extracted, and the abnormal power consumption area is integrated into a high correlation area group.

5. The intelligent dynamic power-saving dispatching method for power grid according to claim 4, characterized in that: The specific process of determining the complementarity between abnormal power consumption areas in the high correlation area group is as follows: In the monitoring period, the monitoring period is divided into several time intervals, the user power consumption of all users in the abnormal power consumption area in the monitoring period is summed to obtain the area power consumption in the monitoring period; If the area power consumption in the monitoring period is higher than the maximum area power consumption, the corresponding monitoring period is recorded as an abnormal peak period, and the area power consumption in the monitoring period is subtracted from the maximum area power consumption to obtain a peak abnormal value. Otherwise, the corresponding monitoring period is recorded as an abnormal low peak period, and the regional power consumption of the monitoring period is subtracted from the regional power consumption minimum value, and the absolute value is taken to obtain a low peak abnormal value; The abnormal peak period, the peak abnormal value, the abnormal low peak period, and the low peak abnormal value are analyzed to determine the peak compensable period and the low peak compensable period, the total number of the peak compensable period and the low peak compensable period of the abnormal power consumption region in the high correlation region group in the monitoring period is counted, and the total number of the abnormal peak period and the abnormal low peak period of the abnormal power consumption region in the high correlation region group is processed by ratio, to obtain a complementarity index, if greater than a complementarity index threshold, it indicates that the complementarity between the abnormal power consumption regions in the high correlation region group is high, otherwise, it indicates that the complementarity is low.

6. The intelligent dynamic power-saving dispatching method for power grid according to claim 5, characterized in that: The determination process of the peak compensable period is as follows: The abnormal peak period of the first abnormal power consumption region in the high correlation region group is integrated into an abnormal peak period sequence in time sequence; The abnormal low peak period of the second abnormal power consumption region in the high correlation region group is integrated into an abnormal low peak period sequence in time sequence; Any abnormal peak period in the abnormal peak period sequence of the first abnormal power consumption region is extracted, and all abnormal low peak periods in the abnormal low peak period sequence of the second abnormal power consumption region are combined one by one to construct a first period analysis group; The peak abnormal value corresponding to the abnormal peak period in the first period analysis group is subtracted from the low peak abnormal value corresponding to the abnormal low peak period, and the absolute value is taken to obtain a first peak-valley abnormal deviation value; If the first peak-valley abnormal deviation value is less than or equal to a peak-valley abnormal deviation standard value, the abnormal peak period in the first period analysis group is recorded as a peak compensable period.

7. The intelligent dynamic power-saving dispatching method for power grid according to claim 6, characterized in that: The determination process of the low peak compensable period is as follows: The abnormal low peak period of the first abnormal power consumption region in the high correlation region group is integrated into an abnormal low peak period sequence in time sequence; The abnormal peak period of the second abnormal power consumption region in the high correlation region group is integrated into an abnormal peak period sequence in time sequence; Any abnormal low peak period in the abnormal low peak period sequence of the first abnormal power consumption region is extracted, and all abnormal peak periods in the abnormal peak period sequence of the second abnormal power consumption region are combined one by one to construct a second period analysis group; The low peak abnormal value corresponding to the abnormal low peak period in the second period analysis group is subtracted from the peak abnormal value corresponding to the abnormal peak period, and the absolute value is taken to obtain a second peak-valley abnormal deviation value; If the second peak-valley abnormal deviation value is less than or equal to a peak-valley abnormal deviation standard value, the abnormal low peak period in the second period analysis group is recorded as a low peak compensable period.

8. The intelligent dynamic power-saving dispatching method for power grid according to claim 1, characterized in that: The specific process of determining the complementarity period characteristics between the abnormal power consumption regions in the compensable region group is as follows: For any abnormal power consumption region (A, B) in a compensable region group, the regional power consumption of the abnormal power consumption region A and the abnormal power consumption region B in the monitoring period is extracted, and is integrated into a regional power consumption sequence A and a regional power consumption sequence B in time sequence, respectively; For the regional electricity consumption sequence, the autocorrelation coefficient of lag k=1, 2, …, k max is calculated , k max is set as the corresponding time length of the monitoring period; Draw the autocorrelation coefficient curve, observe whether there is a peak at a certain lag, and the peak exceeds the preset confidence interval, then take the corresponding k as the best offset amount of the abnormal electricity area (A, B) in the complementary region group; According to the best offset amount, determine the complementary overlap period and the electricity difference value; Take the best offset amount, the complementary overlap period and the electricity difference value of the abnormal electricity area A and the abnormal electricity area B in the complementary region group as the complementary period characteristics.

9. The intelligent dynamic power-saving dispatching method for power grid according to claim 8, characterized in that: The determination process of the complementary overlap period and the electricity difference value is: Statistically determine the overlap period of the peak complementary period and the low peak complementary period of the abnormal electricity area A and the abnormal electricity area B in the complementary region group, and record it as the complementary overlap period of the complementary region group; Take the difference between the peak abnormal value and the low peak abnormal value of the complementary overlap period in the complementary region group, and take the absolute value to obtain the electricity difference value.

10. A power grid intelligent dynamic power saving dispatching system, characterized in that, The system is used to execute the method in any one of claims 1-9, and the system comprises: A data acquisition and analysis module: identifying the abnormal electricity area in the power supply area through the user electricity data in the power supply area; A region correlation analysis module: analyzing the spatial correlation between the abnormal electricity area and the adjacent abnormal electricity area, evaluating the abnormal electricity correlation degree of the abnormal electricity area and the adjacent abnormal electricity area in the geographical space, and determining the high correlation region group; A complementarity analysis module: analyzing the electricity complementarity between the abnormal electricity areas in the high correlation region group, and judging the electricity complementarity between the abnormal electricity areas in the high correlation region group; A complementary overlap period prediction module: if the electricity complementarity between the abnormal electricity areas in the high correlation region group is high, it is recorded as a complementary region group, the complementary period characteristics between the abnormal electricity areas in the complementary region group are determined through the time offset analysis of the abnormal electricity areas in the complementary region group, and a complementary period prediction model is constructed, the complementary period characteristics of the complementary region group are taken as the input of the complementary period prediction model, and the predicted complementary overlap period of the complementary region group is output.

Images