Operation Identification Method and Device for User-Side Energy Storage Device

Abstract

The present invention relates to the technical field of energy management, and specifically provides an operation identification method and device for a user-side energy storage device. The method includes: obtaining the original data of a target enterprise and a control enterprise; wherein, the control enterprise has put an energy storage device into operation. Performing data processing on the original data to obtain the to-be-verified data of the target enterprise and the labeled data of the control enterprise. Inputting the labeled data into a clustering algorithm to output the clustering results before and after the energy storage device is put into operation. Obtaining the identification thresholds before and after the control enterprise puts the energy storage device into operation according to the clustering results. Identifying whether the target enterprise has put the energy storage device into operation according to the identification thresholds and the to-be-verified data. The present invention discovers the inherent structural characteristics of data unsupervised by introducing a clustering algorithm, providing support for the operation research of enterprise energy storage devices. It overcomes the limitation of data disclosure on energy storage operation identification, has good accuracy, and provides a basis for energy management and optimal scheduling.

Description

Technical Field

[0001] The present invention relates to the technical field of energy management, and particularly to a method and device for identifying the operation of a user-side energy storage device. Background Art

[0002] In the power system, the rational utilization of energy storage devices plays a key role in the balance and stable operation of power loads. Energy storage devices can achieve peak shaving and valley filling by storing and releasing electric power resources, thereby improving energy utilization efficiency.

[0003] User-side energy storage is divided into household energy storage and industrial and commercial energy storage according to different terminal users. As industrial and commercial users, enterprises are an important entity of user-side energy storage. Users such as industrial parks, communication base stations, shopping malls, and hospitals can all achieve dynamic capacity increase through the operation of energy storage devices.

[0004] Due to the limited publicly available data disclosed by enterprise users, the operation status of enterprise energy storage devices is hidden in the overall electricity consumption data and cannot be judged; in the prior art, there is a lack of accurate identification of the actual operation behavior of enterprise user energy storage devices, which affects the optimal dispatching of energy and the comprehensive evaluation of energy storage benefits. Summary of the Invention

[0005] The method and device for identifying the operation of a user-side energy storage device provided by the embodiments of the present invention at least solve the problem that due to limited data disclosure, it is impossible to accurately identify whether the energy storage device is put into operation.

[0006] In a first aspect, an embodiment of the present invention provides a method for identifying the operation of a user-side energy storage device, the method including:

[0007] Obtain the original data of a target enterprise and a control enterprise; wherein, the control enterprise has put an energy storage device into operation;

[0008] Perform data processing on the original data to obtain the data to be verified of the target enterprise and the labeled data of the control enterprise;

[0009] Input the labeled data into a clustering algorithm, and output the clustering results before and after the energy storage device is put into operation;

[0010] Obtain the identification thresholds before and after the control enterprise puts the energy storage device into operation according to the clustering results;

[0011] Identify whether the target enterprise has put the energy storage device into operation according to the identification threshold and the data to be verified.

[0012] The method for identifying the operation of a user-side energy storage device provided by the embodiments of the present invention, obtaining the original data of a target enterprise and a control enterprise, includes:

[0013] Obtain the moment when the reference enterprise puts the energy storage device into operation;

[0014] Obtain the first power consumption data of the reference enterprise; wherein, the time period corresponding to the first power consumption data includes the moment when the reference enterprise puts the energy storage device into operation;

[0015] Obtain the second power consumption data of the target enterprise; wherein, the time period corresponding to the second power consumption data includes the target moment to be identified.

[0016] The operation recognition method of the user-side energy storage device provided by the embodiment of the present invention creates data processing on the original data to obtain the data to be verified of the target enterprise and the labeled data of the reference enterprise, including:

[0017] Perform data cleaning on the first power consumption data and the second power consumption data;

[0018] Assign labels before and after the moment when the energy storage device is put into operation, and label the result after cleaning the first power consumption data to obtain the labeled data;

[0019] Assign labels before and after the target moment, and label the result after cleaning the second power consumption data to obtain the data to be verified.

[0020] The operation recognition method of the user-side energy storage device provided by the embodiment of the present invention creates data cleaning on the first power consumption data and the second power consumption data, including:

[0021] Merge the first power consumption data into a first data file, and merge the second power consumption data into a second data file;

[0022] Detect and remove abnormal data in the first data file by the quartile method, and detect and remove abnormal data in the second data file by the quartile method;

[0023] In the first data file, fill the data at the previous moment of the abnormal data to the position of the abnormal data;

[0024] In the second data file, fill the data at the previous moment of the abnormal data to the position of the abnormal data.

[0025] The operation recognition method of the user-side energy storage device provided by the embodiment of the present invention creates inputs the labeled data into a clustering algorithm and outputs the clustering results before and after the energy storage device is put into operation, including:

[0026] Divide the labeled data into multiple time series, and input the time series into the clustering algorithm;

[0027] Initialize the clusters to which the time series belong, and initialize the central time series of each cluster;

[0028] Iteratively update the clustering partition of the time series;

[0029] When the clustering is stable or the maximum number of iterations is reached, output the clustering results before and after the energy storage device is put into operation.

[0030] The operation recognition method of the user-side energy storage device provided by the embodiment of the present invention creatively iteratively updates the clustering partition of the time series, including:

[0031] Calculate and update the central time series of the cluster according to the time series in each cluster under the current clustering partition;

[0032] Calculate the distance between each time series and each updated central time series;

[0033] According to the distance calculation result, assign the time series to the cluster where the nearest central time series is located;

[0034] Update the clustering partition according to the assignment result of the time series.

[0035] The operation recognition method of the user-side energy storage device provided by the embodiment of the present invention creatively calculates and updates the central time series of the cluster according to the time series in each cluster under the current clustering partition, including:

[0036] Multiply the transpose of the time series matrix in the cluster by the time series matrix to obtain the autocorrelation matrix of the time series matrix;

[0037] Obtain the centering matrix according to the length of the time series;

[0038] Multiply the transpose of the centering matrix, the autocorrelation matrix, and the centering matrix to obtain the covariance matrix after centering processing;

[0039] Perform eigenvalue decomposition on the covariance matrix to obtain the maximum eigenvalue of the covariance matrix, and the eigenvector corresponding to the maximum eigenvalue is the updated central time series.

[0040] The operation recognition method of the user-side energy storage device provided by the embodiment of the present invention creatively calculates the distance between each time series and each updated central time series, including:

[0041] Calculate the cross-correlation sequence between the time series and the updated central time series;

[0042] Normalize the cross-correlation sequence;

[0043] Calculate the maximum cross - correlation value, and obtain the distance between the time series and the updated central time series according to the maximum cross - correlation value.

[0044] The operation recognition method of the user - side energy storage device provided by the embodiment of the present invention creates and obtains the recognition threshold before and after the energy storage device is put into operation for the control enterprise according to the clustering result, including:

[0045] Equidistantly divide the time series of the labeled data into several data points;

[0046] Divide the time series of the labeled data into a charging period and a discharging period according to the time - of - use electricity price;

[0047] According to the clustering result before the energy storage device is put into operation, calculate the average power consumption of the data points during the charging period and the average power consumption of the data points during the discharging period;

[0048] According to the clustering result after the energy storage device is put into operation, calculate the average power consumption of the data points during the charging period and the average power consumption of the data points during the discharging period;

[0049] According to the difference in the average power consumption before and after the energy storage device is put into operation during the charging period and the difference in the average power consumption before and after the energy storage device is put into operation during the discharging period, obtain the electricity consumption difference of the control enterprise;

[0050] Calculate the average value of the electricity consumption difference to obtain the recognition threshold; where there are multiple control enterprises, and each control enterprise has one electricity consumption difference.

[0051] The operation recognition method of the user - side energy storage device provided by the embodiment of the present invention creates and identifies whether the target enterprise has put the energy storage device into operation according to the recognition threshold and the data to be verified, including:

[0052] Take the difference between the power consumption after the target time and the power consumption before the target time of the data to be verified as the value to be verified of the target enterprise;

[0053] When the value to be verified is greater than or equal to the recognition threshold, the target enterprise has put the energy storage device into operation;

[0054] When the value to be verified is less than the recognition threshold, the target enterprise has not put the energy storage device into operation at the target time.

[0055] In a second aspect, the present invention also provides an electronic device, including a processor and a memory for storing a program, where the program includes instructions, and when the instructions are executed by the processor, the processor executes the operation recognition method of the user - side energy storage device in any of the above - mentioned embodiments.

[0056] The operation recognition method and device for the user-side energy storage device provided by the embodiment of the present invention create support for the operation research of the enterprise energy storage device by obtaining the electricity consumption data of the enterprises with the energy storage device put into use and introducing the clustering algorithm to unsupervised discover the inherent structural characteristics of the data. It overcomes the limitation of the restricted data disclosure on the energy storage operation recognition, has good accuracy, and provides a basis for energy management and optimal scheduling. Brief Description of the Drawings

[0057] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention, and for those of ordinary skill in the art, other embodiments can be obtained based on these drawings without creative efforts.

[0058] Figure 1 It is a flowchart of the operation recognition method for the user-side energy storage device in the embodiment of the present invention.

[0059] Figure 2 It is a schematic diagram of the clustering result with data labels of a certain target enterprise in the embodiment of the present invention. Detailed Embodiment

[0060] The following will describe the embodiments of the present invention in more detail with reference to the drawings. Although some embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as limited to the embodiments described herein. On the contrary, these embodiments are provided to more thoroughly and completely understand the present invention. It should be understood that the drawings and embodiments of the present invention are only for exemplary purposes and are not used to limit the protection scope of the present invention.

[0061] With the intensification of the global energy crisis and the increasingly serious problem of climate change, the transformation and optimization of the energy structure have become the focus of current research. Energy management is an important path to achieve economic and green development. In the power system, the energy storage device is reasonably used to store the surplus electric energy during the low electricity consumption period and release it during the high electricity consumption period; by cutting peaks and filling valleys to balance the power supply and demand, the grid operation becomes more stable, and at the same time, the consumption capacity of renewable energy is improved, providing support for the economy and flexibility of the grid operation.

[0062] According to the different types of applications of the energy storage device, it can be divided into power generation side energy storage, grid side energy storage and user side energy storage. Among them, the user side energy storage refers to the energy storage device installed at the end customers of the power system, which is independently managed by the user and used to reduce the user's electricity cost.

[0063] The user types of the user-side energy storage devices include industrial and commercial users such as factories, enterprises, shopping malls, hotels, and office buildings, public utility users such as hospitals and schools, and residential users. Due to the generally large electricity consumption load and high electricity price costs of industrial and commercial users, in recent years, the popularity rate of user-side energy storage devices represented by industrial and commercial users has been greatly improved, and users are increasingly introducing energy storage devices.

[0064] After an enterprise introduces an energy storage device, although there is a record of the installation of the energy storage device, it is unknown whether it is actually put into operation. The operation status and operation data are not publicly available, and the operation status of the energy storage device is hidden in the overall power data of the enterprise. Due to limited publicly available data, the power supply company or the grid side cannot accurately understand whether the enterprise has put the energy storage device into actual operation, and there is insufficient research on the identification of the operation behavior of enterprise-level user energy storage devices, which brings inconvenience to work such as energy dispatching and energy storage benefit evaluation.

[0065] Accordingly, with reference to Figure 1 As shown, this embodiment provides an operation identification method for user-side energy storage devices. By introducing a clustering algorithm and combining the general rules of the operation of energy storage devices, the operation identification of energy storage devices is realized by analyzing the change in load characteristics before and after the operation of the energy storage device, providing an important basis for optimizing energy dispatching and evaluating energy storage benefits.

[0066] Specifically, the operation identification method for user-side energy storage devices includes the following steps:

[0067] Step S100, obtain the original data of the target enterprise and the reference enterprise.

[0068] Among them, both the target enterprise and the reference enterprise introduce energy storage devices. The operation status of the energy storage device of the target enterprise is unknown, and the reference enterprise is determined to have put the energy storage device into operation. The method provided in this embodiment mines the internal structural characteristics of the data through the power consumption data of the reference enterprise, providing a basis and support for the operation identification of the energy storage device of the target enterprise.

[0069] As an implementable manner, step S100 includes the following steps:

[0070] Step S110, obtain the moment when the reference enterprise puts the energy storage device into operation.

[0071] Step S120, obtain the first power consumption data of the reference enterprise; among them, the time period corresponding to the first power consumption data includes the moment when the reference enterprise puts the energy storage device into operation.

[0072] Step S130, obtain the second power consumption data of the target enterprise; among them, the time period corresponding to the second power consumption data includes the target moment to be identified.

[0073] The operation recognition method provided in this embodiment is used to recognize whether there is a behavior of an energy storage device being put into operation before and after a target time for a target enterprise. Therefore, the original data of the target enterprise needs to cover the target time to be recognized. Correspondingly, the original data of the control enterprise also needs to include the power consumption data before and after the energy storage device is put into operation.

[0074] Enterprise power consumption data includes various types of power consumption information such as power consumption, power consumption time, and power consumption power. In this embodiment, preferably, both the first power consumption power data and the second power consumption power data in the original data are the total power consumption of enterprise users that can be obtained and are public.

[0075] Step S200: Perform data processing on the original data to obtain the data to be verified of the target enterprise and the labeled data of the control enterprise.

[0076] As an implementable manner, step S200 includes the following steps:

[0077] Step S210: Perform data cleaning on the first power consumption power data and the second power consumption power data of the original data respectively.

[0078] In this embodiment, the data cleaning specifically includes the following steps:

[0079] Step S211: If the first power consumption power data includes multiple data files, merge the multiple data files into one first data file for subsequent processing.

[0080] If the second power consumption power data includes multiple data files, merge the multiple data files into one second data file for subsequent processing.

[0081] Step S212: Detect and remove abnormal data in the first data file through the quartile method; detect and remove abnormal data in the second data file through the quartile method.

[0082] The quartile method identifies and removes outliers in the data based on the interquartile range. The quartile method does not require the data to follow a specific distribution and has good robustness.

[0083] Specifically, sort the data in the first data file according to size. After sorting, calculate the first quartile , the third quartile and the interquartile range , and determine the upper and lower limits of the normal data.

[0084] Among them, the first quartile is the value at the 25% position in the data. If affected by the data volume, the 25% position is not an integer, then take the average of the front and back data as the first quartile . The third quartile It is the value at the 75% position in the data. If affected by the data volume and the 75% position is not an integer, the average of the adjacent data is taken as the third quartile. Interquartile range According to the first quartile and the third quartile calculate. The interquartile range and the upper and lower limits of the normal data are calculated with reference to the following formula:

[0085] ;

[0086] ;

[0087] ;

[0088] In the formula, is the upper limit of the normal data, and the data exceeding the upper limit is excluded as abnormal data; is the lower limit of the normal data, and the data below the lower limit is excluded as abnormal data. For the second data file, the steps of excluding abnormal data by the quartile method are the same as those in the first data file and will not be elaborated here.

[0089] Step S213, in the first data file and the second data file after excluding the abnormal data, fill the positions where the abnormal data is located by the forward filling strategy.

[0090] The forward filling strategy is to fill the position of the abnormal data with the data at the previous moment of the abnormal data. Fill with reference to the following formula:

[0091] ;

[0092] In the formula, is the data at moment. When the data is missing due to being excluded, use the data at the previous moment to fill to position.

[0093] Step S220, label the data after cleaning in step S210.

[0094] Allocate labels before and after the moment when the energy storage device is put into operation, and label the result of the cleaned data of the control enterprise to obtain labeled data. The labeled data has labels before or after being put into operation.

[0095] Assign tags before and after the target time, and label the result of cleaning the target enterprise data to obtain the data to be verified. In this embodiment, by the degree of proximity or similarity between the association between the data to be verified with the pre-target-time tag and the post-target-time tag and the association between the labeled data with the pre-operation tag and the post-operation tag, it is verified whether there is a behavior of the energy storage device being put into operation before and after the target time.

[0096] Step S300: Input the labeled data obtained in step S200 into a clustering algorithm, and output the clustering results before and after the energy storage device is put into operation.

[0097] As an implementable manner, step S300 includes the following steps:

[0098] Step S310: Divide the labeled data into multiple time series according to a certain time scale, and input the time series into the clustering algorithm.

[0099] Step S310 determines the number and length of the time series. In this embodiment, the time period corresponding to the first data file can reach one month or several months. Corresponding to the time period length of the first data file, the time series can be divided with a scale of 24 hours, and a total of N time series are obtained as the input of the clustering algorithm, and the length of each time series is 24 hours. In other embodiments, the division of the time series can be according to other scales, not limited to this.

[0100] In this embodiment, the clustering algorithm is set as the k-shape algorithm, and the k-shape algorithm is improved based on the idea of k-means clustering according to the characteristics of time series data. The k-shape algorithm uses a shape-based distance metric to capture the shape features of time series, and can better reflect the true similarity between time series.

[0101] Similar to the k-means clustering algorithm, the k-shape algorithm needs to specify the number of clusters K in advance. In this embodiment, the number of clusters K = 2, that is to say, the clustering results output by the clustering algorithm include two clusters, corresponding to before and after the energy storage device is put into operation respectively.

[0102] Step S320: Initialize the clusters to which the time series belong, and initialize the central time series of each cluster.

[0103] Randomly initialize the N time series input into the clustering algorithm to one of the clusters. Initialize all data points of the central time series of the two clusters to 0.

[0104] Step 330: Iteratively update the clustering division of the time series.

[0105] Specifically, step S330 includes:

[0106] Step S331: Calculate and update the central time series of each cluster under the current clustering partition. For the first iteration, the current clustering partition is the initialization result in Step S320. Calculate based on the central time series of the current cluster with reference to the following formula:

[0107] ;

[0108] ;

[0109] ;

[0110] .

[0111] In the formula, represents a matrix composed of time series with a length of is the number of time series in the calculated cluster, ≤N. Therefore, when the clustering partition changes, the time series and the number of time series in each cluster may change. When the time series or the number of time series changes, the central time series of the cluster also changes accordingly.

[0112] Multiply the transpose of the time series matrix in the cluster by the time series matrix to obtain the autocorrelation matrix of the time series matrix ; The autocorrelation matrix is used to reflect the similarity between the time series in the cluster.

[0113] Next, obtain the centering matrix according to the length of the time series; In the formula represents the identity matrix, represents the all-ones matrix, and all elements in the matrix are 1. The centering matrix removes the mean component of the time series in the cluster.

[0114] Taking into account the similarity between the time series and the centering adjustment, multiply the transpose of the centering matrix, the autocorrelation matrix , and the centering matrix to obtain the covariance matrix 。

[0115] After that, based on the covariance matrix perform eigenvalue decomposition to obtain the eigenvalues of the covariance matrix 。Take the eigenvector corresponding to the largest eigenvalue as the new central time series of the calculated cluster 。

[0116] For another cluster, when updating the central time series, use the time series in the other cluster as the time series matrix, and the calculation process is the same.

[0117] Step S332, calculate the distance between each time series and each updated central time series in Step S331.

[0118] Taking the central time series of one of the clusters updated in Step S331 as an example, the distance calculation refers to the following formula:

[0119] ;

[0120] In the formula, represents the th time series in the cluster; represents the length of the cross-correlation sequence between the time series and the central time series . The above formula is used to calculate the distance between the time series and the central time series . The distance calculation between the time series and the central time series of another cluster is the same. is used to return the length of the time series . represents the value that is the closest to the input in the direction larger than the input and is a power of 2. For example, for the input of , the function will find an integer q such that ≥ , and < a, and return the integer q.

[0121] ;

[0122] ;

[0123] Subsequently, in the formula represents the cross-correlation sequence between the time series and the central time series . represents the fast Fourier transform, represents the inverse fast Fourier transform. For the cross-correlation sequence Perform normalization processing, indicating the cross-correlation sequence after normalization.

[0124] ;

[0125] Next, calculate the maximum cross-correlation value , and based on the maximum cross-correlation value obtain the time series and the distance from the updated central time series .

[0126] Step S333: According to the distance calculated in step S332, assign the time series to the cluster where the central time series with the closest distance is located. That is to say, for the time series , calculate its distances from the updated central time series of the two clusters respectively, and take the central time series corresponding to the minimum value, indicating that the time series is the closest to this central time series, and then assign the time series to the cluster where this central time series is located.

[0127] Step S334: Perform the calculations and assignments in step S333 for all N time series, and update the clustering partition according to the assignment results of the N time series. It should be noted that when the cluster to which the time series belongs changes, the central time series of the corresponding cluster may also change.

[0128] Based on the updated clustering partition in step S334, repeat steps S331 to S334 in a loop for iteration, continuously update the clustering until the loop process meets the iteration termination condition.

[0129] Step 340: Stop the iteration when the clustering is stable or the maximum number of iterations is reached, and output the clustering results before and after the energy storage device is put into operation.

[0130] Take the clustering being stable or reaching the maximum number of iterations as the iteration termination condition. Among them, when the cluster memberships of all N time series do not change between two adjacent iterations, the clustering is considered stable. In the case of stable clustering, since the cluster memberships of the time series do not change compared to the previous iteration, the corresponding central time series and distance calculation results do not change either. Therefore, continuing the iteration will not output other clustering results, and the clustering is considered stable. The maximum number of iterations is preset, and when the preset threshold of the number of iterations is reached, the iteration is also stopped, and the clustering results are output. The clustering results include two clusters before and after the energy storage device is put into operation, and there are several time series in each cluster.

[0131] Step S400: Obtain the recognition thresholds before and after the control enterprise puts the energy storage device into operation according to the clustering results output in step S300.

[0132] As an implementable manner, step S400 includes the following steps:

[0133] Step S410: Equally divide the time series of the labeled data into several data points.

[0134] In this embodiment, the labeled data is divided into a 24-hour scale for time alignment, and the labeled data records the electricity consumption of the control enterprise with a 15-minute interval as a data point. In other embodiments, the division scale of the time series and the division of data points are not limited to this.

[0135] Step S420: Divide the time series of the labeled data into a charging period and a discharging period according to the stepped electricity price.

[0136] In this embodiment, obtain the stepped electricity price corresponding to each period within a 24-hour time scale in the region where the enterprise is located. As shown in Table 1:

[0137] Table 1

[0138]

[0139] Enterprise energy storage devices usually charge during the valley period with a lower electricity price and discharge during the peak period with a higher electricity price to reduce the electricity cost through peak-valley arbitrage. Through research by the inventor of this embodiment, it is found that enterprise energy storage devices usually complete two charges and two discharges according to the stepped electricity price within 24 hours, and there is no significant difference in the electricity consumption changes during the two charges and discharges. Therefore, in this embodiment, the period from 00:00 to 08:00 (excluding 00:00) is set as the charging period, and the period from 08:00 to 11:00 (excluding 08:00) is set as the discharging period to calculate the electricity consumption difference before and after the energy storage device is put into operation. In other embodiments, it can also be calculated based on the charging period from 11:00 to 17:00 and the discharging period from 17:00 to 22:00.

[0140] Step S430: Calculate the average electricity consumption power of the data points during the charging period and the average electricity consumption power of the data points during the discharging period according to the clustering results before the energy storage device is put into operation.

[0141] Calculate the average electricity consumption power of the data points during the charging period and the average electricity consumption power of the data points during the discharging period according to the clustering results after the energy storage device is put into operation.

[0142] In this embodiment, if a data point is divided every 15 minutes, the charging period includes 32 data points, and the discharging period includes 12 data points. The average electricity consumption power is calculated within the range of these data points. The specific calculation process refers to the following formula:

[0143] ;

[0144] ;

[0145] .

[0146] Wherein, represents the average value of the th group of data within the time period ; represents the data points belonging to the th group within the time period ; represents the data points belonging to the th group within the time period ; takes a value of 1 or 2. The first group represents before the energy storage device is put into operation, and the second group represents after the energy storage device is put into operation. takes a value of 1 or 2. Time period 1 represents the charging period, and time period 2 represents the discharging period.

[0147] Step S440: Obtain the electricity consumption difference of the reference enterprise according to the difference between the average electricity consumption power before and after the energy storage device is put into operation during the charging period, and the difference between the average electricity consumption power before and after the energy storage device is put into operation during the discharging period.

[0148] Specifically, the difference between the average electricity consumption power before and after the energy storage device is put into operation during the charging period is calculated with reference to the following formula:

[0149] ;

[0150] Wherein, represents the difference between the average electricity consumption power after and before the energy storage operation during time period 1, i.e., the charging period.

[0151] The difference between the average electricity consumption power before and after the energy storage device is put into operation during the discharging period is calculated with reference to the following formula:

[0152] ;

[0153] Wherein, represents the difference between the average electricity consumption power after and before the energy storage operation during time period 2, i.e., the discharging period.

[0154] The electricity consumption difference of the reference enterprise is calculated with reference to the following formula:

[0155] .

[0156] Step S450: Calculate the electricity consumption difference The mean value is used to obtain the recognition threshold . Among them, there are multiple reference enterprises, and each reference enterprise has an electricity consumption difference . The recognition threshold is calculated with reference to the following formula:

[0157] .

[0158] In the formula, represents the total number of reference enterprises, represents the electricity consumption difference of the th reference enterprise

[0159] Step S500: Based on the recognition threshold and the data to be verified, identify whether the target enterprise has put the energy storage device into operation

[0160] In some embodiments, the recognition threshold is determined according to the calculation result of step S400. In other embodiments, the recognition threshold can also be appropriately scaled and then compared with the data to be verified

[0161] The difference between the electricity consumption power of the data to be verified after the target time and before the target time is used as the value to be verified of the target enterprise; for example, when identifying whether there is an energy storage device put into operation within 24 hours from 00:00 to 24:00 of the target enterprise, 00:00 - 08:00 is taken as the possible charging stage according to the time-of-use electricity price, 08:00 - 11:00 is taken as the possible discharging stage for verification, the 24 hours where the target time is located is used as the time period to be verified, the difference before and after the target time is taken as the value to be verified, and it is verified whether the energy storage device is put into operation within the time period to be verified

[0162] When the value to be verified is greater than or equal to the recognition threshold , the recognition result is that the target enterprise has put the energy storage device into operation before and after the target time, and the behavior of putting the energy storage device into operation occurs within the time period to be verified

[0163] When the value to be verified is less than the recognition threshold , the recognition result is that the target enterprise has not put the energy storage device into operation at the target time, and there is no behavior of putting it into operation

[0164] Referring to Table 2, in one implementation manner of this embodiment, the average electricity consumption power data of the reference enterprises and the calculation of the electricity consumption difference are shown. In this embodiment, the data shown in Table 2 is taken as an example to specifically illustrate steps S400 to S500:

[0165] Table 2

[0166]

[0167] According to the general law of energy storage operation, during the charging period, due to the charging consumption of the energy storage device, after it is put into operation, the total power consumption of the enterprise increases compared with that before it is put into operation. The average value of each data point before the energy storage device is put into operation during the charging period is 0.0206; the average value of each data point after the energy storage device is put into operation during the charging period is 0.0424; the difference in the average power consumption before and after the energy storage operation during the charging period is 0.02175. The difference is a positive number, and the larger the difference is, the greater the change in power consumption before and after the operation of the energy storage device.

[0168] During the discharging period, due to the power supply of the energy storage device, after it is put into operation, the total power consumption of the enterprise decreases compared with that before it is put into operation. The average value of each data point before the energy storage device is put into operation during the discharging period is 0.1113; the average value of each data point after the energy storage device is put into operation during the discharging period is 0.0540; the difference in the average power consumption before and after the energy storage operation during the discharging period is -0.0574. The difference is a negative number, and the smaller the difference is, the greater the change in power consumption before and after the operation of the energy storage device.

[0169] Subsequently, for this reference enterprise, the power consumption difference is 0.07910. The larger the power consumption difference is, the more obvious the change in the power consumption behavior of this reference enterprise before and after the energy storage device is put into operation. In this embodiment, the power consumption differences of multiple reference enterprises are averaged to obtain the approximate change level of the enterprise's power consumption behavior caused by the energy storage operation, and the finally obtained identification threshold is 0.0625, while the threshold calculated for the target enterprise for identification is 0.0801. Therefore, it is determined that the target enterprise has the behavior of putting the energy storage device into operation.

[0170] This embodiment also clusters the data of the target enterprise to verify the accuracy of the method provided in this embodiment, and the results are shown in Figure 2 the figure. Among them, different color backgrounds are used to distinguish different electricity price periods. The part with a blue background is the valley period of the electricity price, which represents the charging period of the energy storage device for the target enterprise. The part with an orange background is the peak period of the electricity price, and the orange background part adjacent to the blue background corresponds to the discharging period of the energy storage device.

[0171] Figure 2 The upper sub-figure in the figure corresponds to the power consumption waveform under the true label. The pink curve represents the daily curve before the energy storage operation, and the blue curve represents the daily curve after the energy storage operation. Figure 2The sub - graph in the lower middle corresponds to the result obtained by the clustering algorithm. The green curve represents the curve that is correctly classified by the algorithm. The pink represents the pre - operation power consumption data that is wrongly predicted by the algorithm as post - operation data. The blue curve represents the post - operation power consumption data that is wrongly predicted by the algorithm as pre - operation data. The accuracy of the algorithm clustering is 0.87.

[0172] Referring to Figure 2 As can be seen from the upper sub - graph, during the charging period, the blue post - operation curve is higher than the pink pre - operation curve, indicating that energy storage charging consumes power and the power consumption becomes larger. During the discharging period, the blue post - operation curve is lower than the pink pre - operation curve, indicating that energy storage discharging supplies power and the power consumption becomes smaller, which is in line with the general law of energy storage operation. The operation recognition method of the user - side energy storage device provided in this embodiment has a correct recognition result for the target enterprise.

[0173] The operation recognition method of the user - side energy storage device provided in this embodiment obtains the power consumption data of enterprises with energy storage devices already put into operation, introduces a clustering algorithm to unsupervisedly discover the internal structural characteristics of the data, excavates the load characteristic changes before and after the operation of the energy storage device, and reasonably designs the algorithm steps in combination with the general law of energy storage operation, providing support for the operation research of enterprise energy storage devices.

[0174] This embodiment can accurately identify the operation behavior of enterprise energy storage devices, help relevant departments monitor the usage of energy storage devices, and is of great significance for energy management and the development of green energy. This embodiment overcomes the limitation of the restricted data disclosure on energy storage operation recognition, has good accuracy, provides a basis for energy management and optimized dispatching, and is of great significance for optimizing energy management and promoting the construction of smart grids.

[0175] This embodiment also provides an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor. The above - mentioned memory stores a computer program that can be executed by the above - mentioned at least one processor. When the computer program is executed by the above - mentioned at least one processor, it is used to make the electronic device execute the above - mentioned operation recognition method of the user - side energy storage device. The electronic device has all the technical effects of the above - mentioned method.

[0176] The electronic device is intended to represent various forms of digital - electronic computer devices, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the invention described and / or claimed herein.

[0177] This embodiment also provides a non-transitory machine-readable medium storing a computer program, wherein the computer program, when executed by a processor of a computer, is used to cause the computer to execute the operation recognition method of the user-side energy storage device described above.

[0178] This embodiment also provides a computer program product, including a computer program, wherein the computer program, when executed by a processor of a computer, is used to cause the computer to execute the operation recognition method of the user-side energy storage device described above.

[0179] The computer program for implementing the method of this embodiment can be written in any combination of one or more programming languages. These computer programs can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing devices, such that when the computer program is executed by the processor or controller, the functions / operations specified in the flowchart and / or block diagram are implemented. The computer program can be executed entirely on the machine, partially on the machine, executed partially on the machine as an independent software package and partially on a remote machine, or executed entirely on a remote machine or server.

[0180] In the context of the embodiments of the present invention, the machine-readable medium can be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. The machine-readable signal medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, or infrared system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of the machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0181] It should be noted that the term "including" and its variations used in the embodiments of the present invention are open-ended, that is, "including but not limited to". The term "based on" is "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". The modifications of "one" and "plural" mentioned in the embodiments of the present invention are illustrative rather than restrictive, and those skilled in the art should understand that unless clearly stated otherwise in the context, it should be understood as "one or more".

[0182] The user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the embodiments of this invention are all information and data that have been authorized by the user or fully authorized by all parties. Moreover, the collection, use, and processing of relevant data need to comply with the relevant laws, regulations, and standards of the relevant countries and regions, and corresponding operation entrances are provided for users to choose to authorize or reject.

[0183] In the method implementation manners provided by the embodiments of this invention, the various steps recorded can be executed in different orders and / or executed in parallel. In addition, the method implementation manners may include additional steps and / or omit the steps shown. The protection scope of this invention is not limited in this regard.

[0184] The term "embodiment" in this specification means that the specific features, structures, or characteristics described in connection with the embodiment may be included in at least one embodiment of this invention. The phrase appears in various positions in the specification does not necessarily mean the same embodiment, nor does it mean being independent or alternative to other embodiments and mutually exclusive. The various embodiments in this specification are all described in a related manner, and the same or similar parts among the various embodiments are referred to each other. In particular, for device, equipment, and system embodiments, since they are basically similar to method embodiments, the description is relatively simple, and the relevant parts refer to the partial description of the method embodiments.

[0185] The above-described embodiments only represent several implementation manners of this invention, and the description is relatively specific and detailed, but it should not be construed as a limitation of the protection scope. It should be noted that for those of ordinary skill in the art, without departing from the inventive concept of this invention, several modifications and improvements can still be made, and these all belong to the protection scope of this invention. Therefore, the protection scope of this invention should be subject to the appended claims.

Claims

1. A method for identifying the operation of a user-side energy storage device, characterized in that, The method includes: Obtaining the original data of the target enterprise and the control enterprise; wherein, the control enterprise has put the energy storage device into operation; the original data is the total power consumption of the enterprise. Performing data processing on the original data to obtain the data to be verified of the target enterprise and the labeled data of the control enterprise. Inputting the labeled data into a clustering algorithm and outputting the clustering results before and after the energy storage device is put into operation. Obtaining the recognition thresholds of the control enterprise before and after the energy storage device is put into operation according to the clustering results, including: Equidistantly dividing the time series of the labeled data into several data points. Dividing the time series of the labeled data into a charging period and a discharging period according to the time-of-use electricity price. Calculating the average power consumption of the data points during the charging period and the average power consumption of the data points during the discharging period according to the clustering results before the energy storage device is put into operation. Obtaining the electricity consumption difference of the control enterprise according to the difference in the average power consumption before and after the energy storage device is put into operation during the charging period and the difference in the average power consumption before and after the energy storage device is put into operation during the discharging period. Calculating the average value of the electricity consumption differences to obtain the recognition threshold; wherein, there are multiple control enterprises, and each control enterprise has one electricity consumption difference. Identifying whether the target enterprise has put the energy storage device into operation according to the recognition threshold and the data to be verified.

2. The operation recognition method of the user-side energy storage device according to claim 1, wherein Obtaining the original data of the target enterprise and the control enterprise, including: Obtaining the moment when the control enterprise puts the energy storage device into operation. Obtaining the first electricity power data of the control enterprise; wherein, the time period corresponding to the first electricity power data includes the moment when the control enterprise puts the energy storage device into operation. Obtaining the second electricity power data of the target enterprise; wherein, the time period corresponding to the second electricity power data includes the target moment to be identified.

3. The operation recognition method of the user-side energy storage device according to claim 2, characterized in that Performing data processing on the original data to obtain the data to be verified of the target enterprise and the labeled data of the control enterprise, including: Performing data cleaning on the first electricity power data and the second electricity power data. Assigning labels before and after the moment when the energy storage device is put into operation, and labeling the result of the cleaned first electricity power data to obtain the labeled data. Assigning labels before and after the target moment, and labeling the result of the cleaned second electricity power data to obtain the data to be verified.

4. The operation recognition method of the user-side energy storage device according to claim 3, wherein Performing data cleaning on the first electricity power data and the second electricity power data, including: Merging the first electricity power data into a first data file, and merging the second electricity power data into a second data file. Detecting and removing abnormal data in the first data file by the quartile method, and detecting and removing abnormal data in the second data file by the quartile method. In the first data file, filling the data of the previous moment of the abnormal data to the position of the abnormal data. In the second data file, filling the data of the previous moment of the abnormal data to the position of the abnormal data.

5. The operation recognition method of the user-side energy storage device according to claim 1, characterized in that Input the labeled data into a clustering algorithm, and output the clustering results before and after the energy storage device is put into operation, including: Divide the labeled data into multiple time series, and input the time series into a clustering algorithm; Initialize the clusters to which the time series belong, and initialize the central time series of each cluster; Iteratively update the clustering division of the time series; When the clustering is stable or the maximum number of iterations is reached, output the clustering results before and after the energy storage device is put into operation.

6. The operation recognition method of the user-side energy storage device according to claim 5, characterized in that, Iteratively update the clustering division of the time series, including: Calculate and update the central time series of each cluster based on the time series in each cluster under the current clustering division; Calculate the distance between each time series and each updated central time series; Based on the distance calculation results, assign the time series to the cluster where the nearest central time series is located; Update the clustering division according to the assignment results of the time series.

7. The operation recognition method of the user-side energy storage device according to claim 6, characterized in that, Calculate and update the central time series of each cluster based on the time series in each cluster under the current clustering division, including: Multiply the transpose of the time series matrix in the cluster by the time series matrix to obtain the autocorrelation matrix of the time series matrix; Obtain the centering matrix according to the length of the time series; Multiply the transpose of the centering matrix, the autocorrelation matrix, and the centering matrix to obtain the covariance matrix after centering processing; Perform eigenvalue decomposition on the covariance matrix to obtain the maximum eigenvalue of the covariance matrix, and the eigenvector corresponding to the maximum eigenvalue is the updated central time series.

8. The operation recognition method of the user-side energy storage device according to claim 6, characterized in that, Calculate the distance between each time series and each updated central time series, including: Calculate the cross-correlation sequence between the time series and the updated central time series; Normalize the cross-correlation sequence; Calculate the maximum cross-correlation value, and obtain the distance between the time series and the updated central time series based on the maximum cross-correlation value.

9. The operation recognition method of the user-side energy storage device according to claim 1, wherein Identify whether the target enterprise has put the energy storage device into operation according to the recognition threshold and the data to be verified, including: Take the difference between the power consumption after the target time and before the target time of the data to be verified as the value to be verified of the target enterprise; When the value to be verified is greater than or equal to the recognition threshold, the target enterprise has put the energy storage device into operation; When the value to be verified is less than the recognition threshold, the target enterprise has not put the energy storage device into operation at the target time.

10. An electronic device, comprising: A processor and a memory storing a program, characterized in that the program includes instructions that, when executed by the processor, cause the processor to execute the method for identifying the operation of the user-side energy storage device according to any one of claims 1 to 9.

Images