Charging load regulation capability confidence coefficient calculation method and device and storage medium

Abstract

The invention relates to a charging load regulation capability confidence coefficient calculation method and device and a storage medium, which are applied to the technical field of load regulation, and comprise the following steps: extracting a typical charging load curve of each charging station through a clustering algorithm, taking the obtained characteristic data such as the charging station load typical curve, the weather factor and the date factor as the input of a LightGBM prediction model to carry out parameter training of the model, obtaining a load prediction model of the charging station, combining a load prediction result with historical load probability statistics, and calculating the confidence coefficient of the load regulation capability of the charging station; according to the method, the load adjustment capability confidence of the charging station is calculated based on the data driving capability in combination with the load prediction result and the historical load probability statistical data, so that the adjustable capacity of the load of the charging station in different time periods in the future is evaluated, and the prediction accuracy is remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid load regulation, and in particular to a charging load regulation capability confidence calculation method, device and storage medium. Background Art

[0002] As the proportion of renewable energy access to the power grid continues to increase, the operation of the power system has been subject to greater uncertainty and volatility. Demand response and load forecasting are indispensable components of modern power system management. Load forecasting provides basic data for the implementation of demand response, while effective demand response improves the accuracy of load forecasting. The two together promote the stability, cost-effectiveness and sustainability of the power system. Demand-side adjustable resources are becoming increasingly important in improving the stability of the power system. As an important dispatchable load resource, the charging load of electric vehicles requires an accurate and effective assessment of their adjustable capacity.

[0003] Existing research methods usually adopt spatiotemporal distribution modeling methods, especially relying on detailed electric vehicle travel data or road network traffic conditions, ignoring the impact of historical load probability on prediction results, resulting in low prediction accuracy. Summary of the invention

[0004] In view of this, the purpose of the present invention is to provide a charging load regulation capability confidence calculation method, device and storage medium to solve the problem in the prior art that the prediction accuracy is low due to reliance on detailed electric vehicle travel data or road network traffic conditions and ignoring the impact of historical load probability on prediction results.

[0005] According to a first aspect of an embodiment of the present invention, a method for calculating a confidence level of a charging load regulation capability is provided, the method comprising:

[0006] Obtain historical load data of each charging station within a preset time period, and pre-process the acquired historical load data;

[0007] The K-means clustering method is used to cluster the historical load data to obtain the typical load curve of the charging station;

[0008] Obtain the meteorological factor characteristics and working holiday time factor characteristics related to the historical load data at the location of each charging station;

[0009] The typical load curve, meteorological factor characteristics and time factor characteristics of the charging station are used as training data to train the pre-built LightGBM model framework to obtain a charging station load prediction model;

[0010] Predicting load forecast values ​​of the charging station at different times using the charging station load forecasting model;

[0011] According to the historical load data of the charging station, the historical minimum load value and the maximum sudden load drop value at different times are obtained;

[0012] Obtain the load slope lower limit value at different times according to the load forecast value at different times, the historical minimum load value and the maximum sudden load drop value;

[0013] The historical load data of the charging station is used as the input value of the kernel density estimation to obtain the probability density distribution of the historical charging station load at different times;

[0014] Determine the probability density value of the load slope lower limit value at the different moments in the probability density distribution;

[0015] Determine the charging station load adjustment value according to the load landslide lower limit value and load forecast value at different times;

[0016] The confidence level of the charging station load regulation capability at different times is obtained based on the probability density value, the charging station load regulation value and the load forecast value at different times.

[0017] Preferably,

[0018] The preprocessing of the acquired historical load data includes:

[0019] Fill missing values, correct outliers and normalize the acquired historical load data;

[0020] The missing value filling includes: filling the missing values ​​using linear interpolation;

[0021] The outlier correction includes: a horizontal processing method and a vertical processing method;

[0022] The horizontal processing method includes: if the difference between the load value at any moment and the load values ​​at the adjacent moments before and after exceeds the preset horizontal threshold, the load value at the moment is judged to be an abnormal value, and the average value of the load values ​​at the adjacent moments before and after the abnormal value is used as the corrected load value;

[0023] The vertical processing method includes: if the difference between the load value at any moment and the load value at the same moment exceeds a preset vertical threshold; then the load value at the moment is judged to be an abnormal value, and the average of the load values ​​at the same moment is used as the corrected load value;

[0024] The normalization process includes: performing normalization process on the historical load data after missing value filling and abnormal value correction by using a linear normalization method.

[0025] Preferably,

[0026] The K-means clustering method is used to cluster the historical load data to obtain the typical load curve of the charging station, including:

[0027] Determine the number of clusters and initialize the cluster center of each cluster;

[0028] The historical load data are collected to obtain a data set, the distance from each sample point in the data set to the cluster center is obtained, and each sample point is assigned to the cluster to which the cluster center with the closest distance belongs;

[0029] Update the sample points in each cluster, calculate the new cluster center of each cluster, replace the original cluster center with the new cluster center, then recalculate the distance from each sample point to the new cluster center and redistribute the sample points. Repeat the above process until the preset number of iterations is reached or the difference between the new cluster center and the original cluster center is less than the preset threshold. Output the final clustering result and obtain the typical load curve of the charging station.

[0030] Preferably, it also includes:

[0031] The load forecast value output by the charging station load forecast model and the corresponding actual value are evaluated for accuracy using the mean absolute percentage error or the mean absolute error. If the mean absolute percentage error or the mean absolute error is less than a preset error threshold, the accuracy evaluation of the charging station load forecast model passes.

[0032] Preferably,

[0033] The method of obtaining the maximum sudden load drop value at different times according to the historical load data of the charging station includes:

[0034] Get the load value at time t, and get the load value at time t+1;

[0035] The load drop value at time t is obtained by subtracting the load value at time t from the load value at time t+1;

[0036] Obtain all sudden load drop values ​​at time t within a preset historical time period, and select the largest one as the maximum sudden load drop value at time t.

[0037] Preferably,

[0038] The method of obtaining the load slope lower limit value at different times according to the load forecast value at different times, the historical minimum load value and the maximum sudden load drop value includes:

[0039] Subtract the maximum sag load value from the load forecast value at the same moment to obtain the load difference value;

[0040] The load difference is compared with the historical minimum load value at the same moment, and the larger one is selected as the load slope lower limit value at that moment.

[0041] According to a second aspect of an embodiment of the present invention, a charging load regulation capability confidence calculation device is provided, the device comprising:

[0042] Historical data acquisition module: used to obtain historical load data of each charging station within a preset time period and pre-process the acquired historical load data;

[0043] Typical load curve acquisition module: used to cluster historical load data using the K-means clustering method to obtain the typical load curve of the charging station;

[0044] Related feature acquisition module: used to obtain the meteorological factor characteristics and working holiday time factor characteristics related to the historical load data at the location of each charging station;

[0045] Load forecasting model acquisition module: used to train the pre-built LightGBM model framework using the typical load curve, meteorological factor characteristics and time factor characteristics of the charging station as training data to obtain a charging station load forecasting model;

[0046] Prediction module: used to predict the load prediction value of the charging station at different times through the charging station load prediction model;

[0047] The sudden drop load value acquisition module is used to obtain the historical minimum load value and the maximum sudden drop load value at different times according to the historical load data of the charging station;

[0048] Load slope lower limit value acquisition module: used to obtain the load slope lower limit value at different times according to the load forecast value at different times, the historical minimum load value and the maximum sudden load drop value;

[0049] Probability density distribution module: used to use the historical load data of charging stations as the input value of kernel density estimation to obtain the probability density distribution of historical charging station loads at different times;

[0050] Probability density value acquisition module: used to determine the probability density value of the load landslide lower limit value at different times in the probability density distribution;

[0051] Adjustment load value acquisition module: used to determine the adjustment load value of the charging station according to the load landslide lower limit value and load prediction value at different times;

[0052] Confidence acquisition module: used to obtain the confidence of the regulation capacity of the charging station load at different times according to the probability density value, the charging station regulation load value and the load prediction value at different times.

[0053] According to a third aspect of an embodiment of the present invention, there is provided a storage medium, wherein the storage medium stores a computer program, and when the computer program is executed by a main controller, each step in the above method is implemented.

[0054] The technical solution provided by the embodiments of the present invention may have the following beneficial effects:

[0055] This application extracts the typical charging load curve of each charging station through a clustering algorithm, and uses the obtained characteristic data such as the typical charging station load curve, weather factors and date factors as the input of the LightGBM prediction model to train the model parameters, thereby obtaining the load prediction model of the charging station, and combining the load prediction results with the historical load probability statistics to calculate the confidence level of the load regulation capability of the charging station; based on data-driven capabilities, this application combines the load prediction results with the historical load probability statistics to calculate the confidence level of the load regulation capability of the charging station, so as to evaluate the adjustable capacity of the charging station load in different time periods in the future, thereby significantly improving the accuracy of the prediction.

[0056] It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0058] Figure 1 is a flow chart of a method for calculating confidence of charging load regulation capability according to an exemplary embodiment;

[0059] Figure 2 is a schematic diagram of a framework principle of a charging load regulation capability confidence calculation method according to another exemplary embodiment;

[0060] Figure 3 is a system schematic diagram of a charging load regulation capability confidence calculation device according to another exemplary embodiment;

[0061] In the attached figure: 1-historical data acquisition module, 2-typical load curve acquisition module, 3-related feature acquisition module, 4-load prediction model acquisition module, 5-prediction module, 6-sudden drop load value acquisition module, 7-load landslide lower limit value acquisition module, 8-probability density distribution module, 9-probability density value acquisition module, 10-adjustment load value acquisition module, 11-confidence acquisition module. DETAILED DESCRIPTION

[0062] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Instead, they are merely examples of devices and methods consistent with some aspects of the present invention as detailed in the appended claims.

[0063] Embodiment 1

[0064] Figure 1 is a flow chart of a method for calculating the confidence level of charging load regulation capability according to an exemplary embodiment. Figure 1 As shown, the method includes:

[0065] S1, obtaining historical load data of each charging station within a preset time period, and preprocessing the obtained historical load data;

[0066] S2, using K-means clustering method to cluster the historical load data and obtain the typical load curve of the charging station;

[0067] S3, obtaining meteorological factor characteristics and working holiday time factor characteristics related to historical load data at the location of each charging station;

[0068] S4, using the typical load curve, meteorological factor characteristics and time factor characteristics of the charging station as training data to train the pre-built LightGBM model framework to obtain a charging station load prediction model;

[0069] S5, predicting the load prediction value of the charging station at different times by using the charging station load prediction model;

[0070] S6, obtaining historical minimum load values ​​and maximum sudden load drop values ​​at different times according to historical load data of the charging station;

[0071] S7, obtaining the load slope lower limit value at different times according to the load forecast value at different times, the historical minimum load value and the maximum sudden load drop value;

[0072] S8, using the historical load data of the charging station as the input value of the kernel density estimation to obtain the probability density distribution of the historical charging station load at different times;

[0073] S9, determining the probability density values ​​of the load slope lower limit values ​​at different times in the probability density distribution;

[0074] S10, determining a charging station load adjustment value according to the load slope lower limit value and load forecast value at different times;

[0075] S11, obtaining the confidence of the regulation capability of the charging station load at different times according to the probability density value at different times, the charging station regulation load value and the load prediction value;

[0076] It is understandable that if Figure 2 As shown, this embodiment collects the historical load data of the charging station with a time scale of 15 minutes; and performs standardization processing on the collected historical load of the charging station, specifically including:

[0077] Missing value filling:

[0078] The linear interpolation method is used to deal with the missing value problem in the historical charging station load data. The formula of the linear interpolation method is as follows:

[0079]

[0080] In the formula, y m is the load value at the mth moment; y m+1 is the load value at the m+1th moment; y m+k is the missing load value at time m+k;

[0081] Outlier correction:

[0082] Horizontal processing method

[0083] If the difference between the current load value and the adjacent load value in the load data set exceeds the preset value, it can be determined that the data is abnormal and the horizontal processing method can be used to handle the abnormality. The processing method is as follows:

[0084]

[0085]

[0086] In the formula, is the threshold set by the horizontal processing method, y(t) is the load value at the tth moment, and y(t) is the corrected load value.

[0087] Vertical processing method

[0088] Since load data has a strong periodicity, if the data at a certain moment differs too much from the data at the same moment and exceeds the set threshold, the value will be marked as abnormal data. The abnormal value will be corrected according to the mean value of the load data at the same moment. The processing method is as follows:

[0089]

[0090]

[0091] In the formula, β is the threshold value set by the vertical processing method, y(t) is the load value at the tth moment, is the average load value of the same type of time in the load data, and y(t) is the corrected load value.

[0092] Normalization:

[0093] In order to ensure that different variables such as load and weather factors do not cause gradient deviation due to different magnitudes, it is necessary to normalize the historical data to obtain the same weight tendency and map different variables to the value range [0,1]. Therefore, this embodiment uses linear normalization to normalize the input variables. The normalization formula is as follows:

[0094]

[0095] In the formula, is the data after load normalization; x is the initial data of load; is the average value of the initial load data; x max is the maximum value of the initial load data; x min is the minimum value of the initial load data;

[0096] The K-means clustering method is used to extract the typical load curve of the charging station, including:

[0097] First, determine the number of clusters and initialize the cluster center of each cluster;

[0098] Secondly, calculate the distance from each sample point in the data set to the cluster center, and assign each sample point to the cluster to which the nearest cluster center belongs;

[0099] Finally, according to each sample point in the updated cluster, a new cluster center is calculated, and the original cluster center is replaced with the new cluster center. Then, the distance from each sample point to the new cluster center is recalculated and the sample points are redistributed. This process is repeated until the cluster center no longer changes significantly or the maximum number of iterations is reached, and the final clustering result, i.e., the typical load curve of the charging station, is output.

[0100] The data of meteorological factors such as temperature, humidity, solar radiation and wind speed near the charging station are collected, and the typical load curve of the charging station, meteorological factors and time characteristic factors (working days and holidays) extracted by clustering are used as the input data of the pre-built LightGBM model framework to form a data set, which is divided into a training set and a test set in a ratio of 7:3 to train the LightGBM model framework, including:

[0101] Initialize and calculate the constant value that minimizes the loss function:

[0102]

[0103] Where L() is the loss function, f0(x) is the initial prediction function, which is usually used as the starting point of the entire gradient boosting decision tree model, where x is the model input; y i is a one-dimensional label; c is the model output, i.e. the predicted value; argmin is a function that obtains the value of the independent variable when a function reaches its minimum value;

[0104] The negative gradient of the loss function is used to estimate the residual r mi , the calculation formula is:

[0105]

[0106] Where f(x) is the prediction function of the input variable x, f(x i ) is the prediction function of the i-th sample, f m-1 (x) is the prediction function at the m-1th iteration.

[0107] Determine the leaf node area as R mj , use linear search to estimate the value of the leaf node area to minimize the loss function, the formula is:

[0108]

[0109] Where: c mj is the constant value of the jth leaf node of the mth regression tree, f m-1 (x i ) is the prediction function of the i-th sample at the m-1th iteration;

[0110] Update the regression tree:

[0111]

[0112] In the formula, f m (x) is the prediction function at the mth iteration; J is the number of leaf nodes; I is the indicator function, which takes a value of 0 or 1, and is used to indicate whether the sample belongs to a certain leaf node area;

[0113] After the iteration is completed, the final result of the LightGBM prediction model is obtained:

[0114]

[0115] In the formula, f m (x) is the sum of the predicted values ​​of all regression trees;

[0116] Prediction result accuracy evaluation:

[0117] The mean absolute percentage error (MAPE) or mean absolute error (MAE) indicator is selected as the evaluation standard of the model prediction error to judge the prediction effect of the prediction model. The expressions of MAPE and MAE are as follows:

[0118]

[0119]

[0120] Where n is the number of predictions; A i is the actual value of the charging station load; Y i is the predicted value of the charging station load;

[0121] When MAPE or MAE is less than the preset error threshold, it indicates that the model's prediction result load target can be used to predict the charging station load value;

[0122] According to the historical charging station load forecast value at different times, the historical minimum load value at the same time, and the maximum sudden load drop value, the lower limit of the load landslide at different times is obtained, where:

[0123] The maximum sudden load drop is obtained based on the historical sudden load drop values, that is, at time t, the sudden load drop = the load value at time t+1 - the load value at time t. The sudden load drop values ​​at time t within a period of time in history are counted, and the maximum sudden load drop value within this period is taken as the maximum sudden load drop value;

[0124] Load slope lower limit = max (the minimum charging station load value at the same time in history, the charging station load prediction value - the maximum sudden drop load value), that is, the difference between the charging station load prediction value and the maximum sudden drop load value is compared with the minimum charging station load value at the same time in history, and the larger one is selected as the load slope lower limit;

[0125] The historical load results of the charging station are used as the input value of the kernel density estimation to calculate the probability density distribution K(x) of the historical charging station load at different times;

[0126] According to the obtained load slope lower limit, determine the corresponding probability density value in the load probability density distribution function

[0127] Select a value between the load slope lower limit and the load forecast value as the charging station load adjustment value.

[0128] Finally, using the probability density value Charging station adjusts load value And the predicted value of charging station load The confidence level of the load regulation capability of the charging station is calculated. The closer the load regulation value of the charging station is to the minimum load value of the charging station, the lower the confidence level. The confidence level calculation formula is as follows:

[0129]

[0130] In the formula, C t is the confidence level of the charging station load regulation capability at time t, The load value of the charging station at time t is adjusted. is the predicted value of the charging station load at time t, It is the probability density value corresponding to the lower limit of the load slope at time t in the load probability density distribution function.

[0131] Embodiment 2

[0132] Figure 3 is a system schematic diagram of a charging load regulation capability confidence calculation device according to another exemplary embodiment, the device comprising:

[0133] Historical data acquisition module 1: used to obtain historical load data of each charging station within a preset time period and pre-process the acquired historical load data;

[0134] Typical load curve acquisition module 2: used to cluster the historical load data using the K-means clustering method to obtain the typical load curve of the charging station;

[0135] Related feature acquisition module 3: used to obtain the meteorological factor characteristics and working holiday time factor characteristics related to the historical load data at the location of each charging station;

[0136] Load forecasting model acquisition module 4: used to train the pre-built LightGBM model framework using the typical load curve, meteorological factor characteristics and time factor characteristics of the charging station as training data to obtain a charging station load forecasting model;

[0137] Prediction module 5: used to predict the load prediction value of the charging station at different times through the charging station load prediction model;

[0138] The sudden drop load value acquisition module 6 is used to obtain the historical minimum load value and the maximum sudden drop load value at different times according to the historical load data of the charging station;

[0139] Load slope lower limit value acquisition module 7: used to acquire the load slope lower limit value at different times according to the load forecast value at different times, the historical minimum load value and the maximum sudden load drop value;

[0140] Probability density distribution module 8: used to use the historical load data of the charging station as the input value of the kernel density estimation to obtain the probability density distribution of the historical charging station load at different times;

[0141] Probability density value acquisition module 9: used to determine the probability density value of the load slope lower limit value at different times in the probability density distribution;

[0142] The load adjustment value acquisition module 10 is used to determine the load adjustment value of the charging station according to the load slope lower limit value and the load prediction value at different times;

[0143] Confidence acquisition module 11: used to acquire the confidence of the regulation capability of the charging station load at different times according to the probability density value, the charging station regulation load value and the load prediction value at different times.

[0144] Embodiment three:

[0145] This embodiment provides a storage medium, wherein the storage medium stores a computer program, and when the computer program is executed by a main controller, each step in the above method is implemented;

[0146] It is understandable that the storage medium mentioned above can be a read-only memory, a magnetic disk or an optical disk, etc.

[0147] It can be understood that the same or similar parts of the above embodiments can be referenced to each other, and the contents not described in detail in some embodiments can refer to the same or similar contents in other embodiments.

[0148] It should be noted that, in the description of the present invention, the terms "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance. In addition, in the description of the present invention, unless otherwise specified, the meaning of "plurality" refers to at least two.

[0149] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code that includes one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present invention includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present invention belong.

[0150] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0151] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0152] In addition, each functional unit in each embodiment of the present invention may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0153] The storage medium mentioned above can be a read-only memory, a magnetic disk or an optical disk, etc.

[0154] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0155] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. A method for calculating the confidence of charging load regulation capability, characterized in that: The method comprises: Obtain historical load data of each charging station within a preset time period, and pre-process the acquired historical load data; The K-means clustering method is used to cluster the historical load data to obtain the typical load curve of the charging station; Obtain the meteorological factor characteristics and working holiday time factor characteristics related to the historical load data at the location of each charging station; The typical load curve, meteorological factor characteristics and time factor characteristics of the charging station are used as training data to train the pre-built LightGBM model framework to obtain a charging station load prediction model; Predicting load forecast values ​​of the charging station at different times using the charging station load forecasting model; According to the historical load data of the charging station, the historical minimum load value and the maximum sudden load drop value at different times are obtained; Obtain the load slope lower limit value at different times according to the load forecast value at different times, the historical minimum load value and the maximum sudden load drop value; The historical load data of the charging station is used as the input value of the kernel density estimation to obtain the probability density distribution of the historical charging station load at different times; Determine the probability density value of the load slope lower limit value at the different moments in the probability density distribution; Determine the charging station load adjustment value according to the load landslide lower limit value and load forecast value at different times; The confidence level of the charging station load regulation capability at different times is obtained based on the probability density value, the charging station load regulation value and the load forecast value at different times.

2. The method according to claim 1, characterized in that The preprocessing of the acquired historical load data includes: Fill missing values, correct outliers and normalize the acquired historical load data; The missing value filling includes: filling the missing values ​​using linear interpolation; The outlier correction includes: a horizontal processing method and a vertical processing method; The horizontal processing method includes: if the difference between the load value at any moment and the load values ​​at the adjacent moments before and after exceeds the preset horizontal threshold, the load value at the moment is judged to be an abnormal value, and the average value of the load values ​​at the adjacent moments before and after the abnormal value is used as the corrected load value; The vertical processing method includes: if the difference between the load value at any moment and the load value at the same moment exceeds a preset vertical threshold; then the load value at the moment is judged to be an abnormal value, and the average of the load values ​​at the same moment is used as the corrected load value; The normalization process includes: performing normalization process on the historical load data after missing value filling and abnormal value correction by using a linear normalization method.

3. The method according to claim 2, characterized in that The K-means clustering method is used to cluster the historical load data to obtain the typical load curve of the charging station, including: Determine the number of clusters and initialize the cluster center of each cluster; The historical load data are collected to obtain a data set, the distance from each sample point in the data set to the cluster center is obtained, and each sample point is assigned to the cluster to which the cluster center with the closest distance belongs; Update the sample points in each cluster, calculate the new cluster center of each cluster, replace the original cluster center with the new cluster center, then recalculate the distance from each sample point to the new cluster center and redistribute the sample points. Repeat the above process until the preset number of iterations is reached or the difference between the new cluster center and the original cluster center is less than the preset threshold. Output the final clustering result and obtain the typical load curve of the charging station.

4. The method according to claim 3, characterized in that: Also includes: The load forecast value output by the charging station load forecast model and the corresponding actual value are evaluated for accuracy using the mean absolute percentage error or the mean absolute error. If the mean absolute percentage error or the mean absolute error is less than a preset error threshold, the accuracy evaluation of the charging station load forecast model passes.

5. The method according to claim 4, characterized in that The method of obtaining the maximum sudden load drop value at different times according to the historical load data of the charging station includes: Get the load value at time t, and get the load value at time t+1; The load drop value at time t is obtained by subtracting the load value at time t from the load value at time t+1; Obtain all sudden load drop values ​​at time t within a preset historical time period, and select the largest one as the maximum sudden load drop value at time t.

6. The method according to claim 5, characterized in that The method of obtaining the load slope lower limit value at different times according to the load forecast value at different times, the historical minimum load value and the maximum sudden load drop value includes: Subtract the maximum sag load value from the load forecast value at the same moment to obtain the load difference value; The load difference is compared with the historical minimum load value at the same moment, and the larger one is selected as the load slope lower limit value at that moment.

7. A charging load regulation capability confidence calculation device, characterized in that: The device comprises: Historical data acquisition module: used to obtain historical load data of each charging station within a preset time period and pre-process the acquired historical load data; Typical load curve acquisition module: used to cluster historical load data using the K-means clustering method to obtain the typical load curve of the charging station; Related feature acquisition module: used to obtain the meteorological factor characteristics and working holiday time factor characteristics related to the historical load data at the location of each charging station; Load forecasting model acquisition module: used to train the pre-built LightGBM model framework using the typical load curve, meteorological factor characteristics and time factor characteristics of the charging station as training data to obtain a charging station load forecasting model; Prediction module: used to predict the load prediction value of the charging station at different times through the charging station load prediction model; The sudden drop load value acquisition module is used to obtain the historical minimum load value and the maximum sudden drop load value at different times according to the historical load data of the charging station; Load slope lower limit value acquisition module: used to obtain the load slope lower limit value at different times according to the load forecast value at different times, the historical minimum load value and the maximum sudden load drop value; Probability density distribution module: used to use the historical load data of charging stations as the input value of kernel density estimation to obtain the probability density distribution of historical charging station loads at different times; Probability density value acquisition module: used to determine the probability density value of the load landslide lower limit value at different times in the probability density distribution; Adjustment load value acquisition module: used to determine the adjustment load value of the charging station according to the load landslide lower limit value and load prediction value at different times; Confidence acquisition module: used to obtain the confidence of the regulation capacity of the charging station load at different times according to the probability density value, the charging station regulation load value and the load prediction value at different times.

8. A storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by the main controller, each step in the charging load regulation capability confidence calculation method as described in any one of claims 1 to 6 is implemented.

Images