Robust-based outlier anomaly power consumption detection result verification method and system

Abstract

The application discloses a kind of outlier abnormal power consumption detection result verification method and system based on robustness, which is based on the metering automation system to obtain the power consumption data of users in the area, and carries out data cleaning and preprocessing;Using screening method and embedding method, the characteristic index related to the line loss of user daily power consumption in the area is analyzed, and heuristic search and factor analysis between characteristics are carried out on the characteristic index to select key features;Based on the key features, an outlier algorithm model is constructed, and model training is carried out to find suitable parameter threshold, and output the suspected user with abnormal power consumption behavior;Data verification is carried out on the case set.The application can maintain stability and reliability when facing data uncertainty, noise and external interference.

Description

Technical Field

[0001] This invention relates to the field of power system technology, and more specifically, to a robust method and system for verifying outlier abnormal power consumption detection results. Background Technology

[0002] With the deepening of electricity market reforms and the increasingly diversified and complex nature of user electricity consumption behavior, analyzing electricity consumption behavior using low-voltage distribution area user data to achieve more refined and effective management of low-voltage distribution areas has become a crucial issue urgently needing to be addressed in current power management and operation. This demand stems from the constantly changing electricity market environment and the personalized and diversified needs of users for electricity use. However, in practice, analyzing electricity consumption behavior using low-voltage distribution area user data faces a series of challenges.

[0003] First, electricity consumption information in low-voltage distribution areas is relatively scarce, mainly due to limitations in data acquisition technology and insufficient data sharing mechanisms. Because of the insufficient data volume, analysts struggle to obtain comprehensive and accurate user electricity consumption data, making it difficult to build effective electricity consumption behavior analysis models.

[0004] Secondly, the sheer number of low-voltage distribution areas increases the complexity and workload of data analysis. Each distribution area has a certain number of users, and each user's electricity consumption behavior is unique, resulting in an extremely large amount of data that analysts need to process. How to efficiently process and analyze this data within limited resources and time has become a pressing problem.

[0005] Furthermore, the factors influencing user electricity consumption and transformer area line losses are complex and multifaceted. These factors include, but are not limited to, weather conditions, equipment condition, and user electricity consumption habits. The interplay and interaction of these factors make electricity consumption behavior analysis more complex and difficult. At the same time, the significant uncertainty surrounding the changes in these factors further increases the difficulty of the analysis.

[0006] Furthermore, the poor robustness of detecting suspected abnormal electricity consumption is a prominent issue in current electricity behavior analysis. Due to the diversity and concealment of abnormal electricity consumption behaviors, as well as the presence of data noise and interference, existing detection methods suffer from high false alarm and false negative rates when identifying abnormal electricity consumption behaviors. This not only reduces the accuracy of the analysis but also increases the difficulty and cost of governance. Summary of the Invention

[0007] To overcome the problem that existing electricity consumption behavior analysis methods cannot effectively analyze abnormal electricity consumption of low-voltage users, this invention provides a robust method and system for verifying outlier abnormal electricity consumption detection results.

[0008] The technical solution of this invention is as follows:

[0009] In a first aspect, the present invention provides a robust method for verifying outlier abnormal power consumption detection results, comprising the following steps:

[0010] Step S1: Obtain electricity consumption data of users in the distribution area based on the metering automation system, and perform data cleaning and preprocessing;

[0011] Step S2: Using screening and embedding methods, analyze the characteristic indicators related to the user's daily electricity consumption and the line loss of the transformer area, and perform heuristic search and factor analysis between the characteristic indicators to select key features.

[0012] Step S3: Construct an outlier algorithm model based on the key features, train the model to find suitable parameter thresholds, and output suspected users with abnormal electricity consumption behavior.

[0013] Step S4: Perform data validation on the case set.

[0014] As a preferred embodiment of the present invention, step S1 specifically includes the following sub-steps:

[0015] Step S101: Remove daily power data with power input to the transformer area ≤ 0, and remove daily power data with line loss rate < -2% or line loss rate > 40%;

[0016] Step S102: Combining the methods of horizontal and vertical comparison of the substation area, interpolation is performed on the missing daily electricity consumption data of the user.

[0017] Step S103: Use confidence intervals to identify and correct abnormal electricity consumption data of users.

[0018] As a preferred embodiment of the present invention, in step S102, when the number of consecutive days of missing daily electricity consumption for a user is ≤7 days, cubic spline interpolation is used to interpolate the missing daily electricity consumption data for the user, and the daily electricity supply of the transformer area after interpolation meets the line loss rate condition in step S101.

[0019] As a preferred embodiment of the present invention, in step S103, the formula for the confidence interval is as follows:

[0020]

[0021] Wherein, mean is the average value of user electricity consumption data, standard deviation is a measure of the dispersion of user electricity consumption data, z is the z-score of the standard normal distribution, and n is the number of data points.

[0022] As a preferred embodiment of the present invention, user electricity consumption data located outside the confidence interval is identified as abnormal electricity consumption data and corrected to the boundary value of the confidence interval, while the power supply to the transformer area is corrected accordingly.

[0023] As a preferred embodiment of the present invention, the feature indicators include daily power consumption, daily power consumption change, line loss power consumption, line loss power consumption change rate, K value, and restored power consumption, wherein the K value is the indicator influence degree, and the restored power consumption is the user's actual power consumption;

[0024] The K value is calculated based on the user's actual electricity consumption on the current day, the user's actual electricity consumption on the previous day, the line loss of the transformer area on the current day, and the line loss of the transformer area on the previous day.

[0025] As a preferred embodiment of the present invention, the formula for calculating the K value based on the user's actual electricity consumption on the current day, the user's actual electricity consumption on the previous day, the line loss of the transformer area on the current day, and the line loss of the transformer area on the previous day is as follows:

[0026]

[0027] As a preferred embodiment of the present invention, the restored charge is obtained by constructing a charge conservation equation and solving it using the least squares method to obtain the user error rate.

[0028] As a preferred embodiment of the present invention, step S3 specifically includes the following sub-steps:

[0029] Step S301: Select sample data of electricity consumption jumps;

[0030] Step S302: Select abnormal power consumption sample data;

[0031] Step S303: Remove sample data with abnormal line loss power.

[0032] Step S304: Based on the K-value sample data obtained from steps S301-S303 and the parameter threshold, determine whether the user is an abnormal suspected user.

[0033] Step S305: Output the list of abnormal users.

[0034] Secondly, the present invention provides a robust outlier anomaly power consumption detection result verification system, comprising:

[0035] The data acquisition and preprocessing module is used to acquire electricity consumption data of users in the distribution area based on the metering automation system, and to perform data cleaning and preprocessing.

[0036] The feature selection module is used to analyze the feature indicators related to the user's daily electricity consumption and the line loss of the transformer area using screening and embedding methods, and to perform heuristic search and factor analysis between the feature indicators to select key features.

[0037] The model building and output module is used to build an outlier algorithm model based on the key features, train the model to find suitable parameter thresholds, and output suspected users with abnormal electricity consumption behavior.

[0038] The data validation module is used to perform data validation on the case set.

[0039] According to the above-described solution, the beneficial effects of this invention are as follows:

[0040] 1. The present invention provides a robust outlier abnormal power consumption detection result verification method, which mines outliers in user power consumption behavior in low-voltage distribution areas, extracts user power consumption behavior characteristics in depth, and mines abnormal power consumption behavior within users. At the same time, it improves robustness by optimizing the algorithm, so that it can still ensure the stability and reliability of detection results when facing various complex situations such as data uncertainty, noise and external interference.

[0041] 2. By introducing the impact of actual user electricity consumption changes on the line loss of the transformer area, an indicator influence degree K is constructed as an abnormal user criterion. The calculation results of the model are objectively verified, and the user status is comprehensively analyzed to improve the identification efficiency, accuracy and robustness, providing an effective method for analyzing abnormal user electricity consumption behavior. Attached Figure Description

[0042] Figure 1 This is a flowchart illustrating the robust outlier anomaly power consumption detection result verification method of the present invention.

[0043] Figure 2 for Figure 1 A detailed flowchart of step S1 is shown below;

[0044] Figure 3 for Figure 1 A detailed flowchart of step S3 is shown below;

[0045] Figure 4 This is a schematic diagram of the robust outlier abnormal power consumption detection result verification system of the present invention. Detailed Implementation

[0046] To make the technical problems to be solved, the technical solutions, and the beneficial effects of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0047] It should be noted that the terms "comprising" and "having," and any variations thereof, in the specification and claims of this invention are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the steps or units listed, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0048] Please refer to Figure 1 This embodiment provides a robust method for verifying outlier abnormal power consumption detection results, including the following steps:

[0049] Step S1: Obtain electricity consumption data of users in the distribution area based on the metering automation system, and perform data cleaning and preprocessing.

[0050] In one embodiment, please refer to Figure 2 Step S1 specifically includes the following sub-steps:

[0051] Step S101: Data showing abnormal power supply and line loss in the transformer area due to factors such as data collection are removed. After removing abnormal data, the remaining data is closer to the actual situation, providing a more accurate basis for subsequent data analysis and processing, ensuring the data quality of subsequent analysis, and avoiding interference from abnormal data on the overall results. Specifically, daily power supply data with power supply ≤ 0 is removed, as are daily power supply data with a line loss rate < -2% or > 40%.

[0052] Step S102: Combining horizontal and vertical comparison methods for different power distribution areas, the missing daily electricity consumption data for users is interpolated to fill in the gaps and make the data more complete. Specifically, when the number of consecutive missing daily electricity consumption days for a user is ≤7 days, cubic spline interpolation is used to interpolate the missing daily electricity consumption data, ensuring that the daily power supply of the power distribution area after interpolation meets the line loss rate condition in step S101, further guaranteeing the accuracy and rationality of the data. Using cubic spline interpolation ensures that the interpolated data maintains the same trend as the original data, thus preserving data continuity.

[0053] Step S103: Abnormal electricity consumption is identified and corrected using confidence intervals to improve the accuracy and reliability of user electricity consumption data. A confidence interval is an interval estimate that provides a range of possible values ​​for a parameter, which has a certain probability (i.e., a confidence level) of containing the true value of the parameter. In this invention, confidence intervals are used to assess the normal fluctuation range of user electricity consumption data. Specifically, the formula for a confidence interval is as follows:

[0054]

[0055] Wherein, the mean is the average value of the user electricity consumption data; the standard deviation is a measure of the dispersion of the user electricity consumption data; z is the z-score of the standard normal distribution, which is set to 1.96 here, indicating a confidence level of 95%, that is, there is a 95% certainty that the true parameter value will fall within the calculated confidence interval; n is the number of data points, that is, the number of user electricity consumption data points used to calculate the mean and standard deviation.

[0056] When using confidence intervals to identify and correct abnormal electricity consumption data, the process begins by first identifying electricity consumption data outside the confidence interval as abnormal. Then, the identified abnormal electricity consumption data is corrected to the boundary values ​​of the confidence interval. If the abnormal electricity consumption data is less than the lower limit of the confidence interval, it is corrected to the lower limit value; if the abnormal electricity consumption data is greater than the upper limit of the confidence interval, it is corrected to the upper limit value. Finally, since user electricity consumption data is interconnected, after correcting the user electricity consumption data, the power supply to the distribution area also needs to be corrected accordingly to maintain data consistency.

[0057] Step S2: Using screening and embedding methods, analyze the characteristic indicators related to the daily electricity consumption of users and the line loss of the transformer area, and perform heuristic search and factor analysis between the characteristic indicators to select key features.

[0058] The advantage of using screening and embedding methods to analyze feature indicators lies in the ability to systematically evaluate the correlation between various features and users' daily electricity consumption and transformer line loss. Screening methods can quickly narrow down the range of features, removing those that are unrelated or weakly correlated with the target variable (such as transformer line loss), while embedding methods can automatically select features that significantly contribute to model performance during model building. Combining screening and embedding methods helps reduce noise and redundancy, improving the efficiency and accuracy of subsequent analyses.

[0059] Heuristic search is an effective feature selection strategy that can find a near-optimal subset of features within a reasonable timeframe. Through heuristic search, features crucial for predicting or analyzing transformer line losses can be further filtered out, thereby improving the model's interpretability and predictive performance.

[0060] Factor analysis of features helps reveal the potential relationships and structures between them. Through factor analysis, multiple related features can be summarized into a few independent factors. These factors can more concisely express the information of the original features, which helps to understand the causes and influencing factors of transformer line loss and provides strong support for subsequent analysis and decision-making.

[0061] In one embodiment, the characteristic indicators include daily electricity consumption, daily electricity consumption change, line loss electricity consumption, line loss electricity consumption change rate, K value, and restored electricity consumption.

[0062] The K-value represents the degree of influence of the indicator. The K-value is calculated based on the actual user electricity consumption on the current day, the actual user electricity consumption on the previous day, and the line loss of the transformer area on the current day and the line loss of the transformer area on the previous day. Specifically, the formula for calculating the K-value based on the actual user electricity consumption on the current day, the actual user electricity consumption on the previous day, and the line loss of the transformer area on the current day and the line loss of the transformer area on the previous day is as follows:

[0063]

[0064] The K-value, by comparing changes in actual user electricity consumption with changes in line loss in the distribution area, can reflect the degree of impact of changes in user electricity consumption on changes in line loss in the distribution area. This helps to more accurately assess the contribution of user behavior to line loss in the distribution area, thereby providing a basis for developing targeted loss reduction measures.

[0065] The restored electricity consumption is the user's actual electricity consumption. This restored electricity consumption is obtained by constructing an electricity conservation equation, solving it using the least squares method to obtain the user's error rate, and then restoring the original consumption. Specifically, the formula for the electricity conservation equation is as follows:

[0066]

[0067] Where y is the total power supply of the transformer substation, and e j Let the error be the value of the electricity meter for the j-th user. Let e ​​be the electricity consumption of the j-th user's electricity meter. y e0 refers to the line loss rate, and e0 refers to the fixed loss.

[0068] Specifically, the formula for restoring the charge is as follows:

[0069]

[0070] The restored electricity consumption data takes into account the impact of factors such as electricity meter error and line loss rate on power supply, thereby improving the accuracy and reliability of electricity consumption data. At the same time, the restored electricity consumption data can more realistically reflect users' electricity consumption behavior, providing strong support for subsequent electricity consumption analysis and forecasting.

[0071] Step S3: Construct an outlier algorithm model based on key features, train the model to find suitable parameter thresholds, and output suspected users with abnormal electricity consumption behavior.

[0072] In one embodiment, please refer to Figure 3 Step S3 specifically includes the following sub-steps:

[0073] Step S301: Select sample data for electricity consumption jumps. Specifically, select sample data where the restored electricity consumption change rate is outside the observed value; select sample data where the daily electricity consumption change is outside the observed value. Observed value = upper / lower quartile ± 1.5 × upper quartile - lower quartile). By selecting sample data where the restored electricity consumption change rate and daily electricity consumption change exceed the observed value, abnormal situations of sudden increases or decreases in electricity consumption can be effectively captured, providing important clues for the identification of abnormal electricity consumption behavior, thereby quickly locating the user group with abnormal electricity consumption behavior.

[0074] Step S302: Select abnormal electricity consumption sample data. Specifically, select sample data where the rate of change in line loss is greater than a threshold (10% or the upper or lower quartile), and the user's electricity consumption is directly proportional to the line loss. By setting a threshold for the rate of change in line loss and combining it with the proportional relationship between user electricity consumption and line loss, it is possible to more accurately identify users whose electricity consumption changes little but whose line loss increases abnormally, thus improving the accuracy of abnormal electricity consumption behavior identification.

[0075] Step S303: Remove sample data with abnormal line loss. Specifically, remove sample data where the rate of change in line loss is within the observed value and the line loss on the previous day was not abnormal. By removing sample data where the rate of change in line loss is within the observed value and the line loss on the previous day was not abnormal, misjudgments caused by accidental factors or data errors can be reduced, which helps improve the model's accuracy in identifying abnormal electricity consumption behavior and reduces unnecessary interference.

[0076] Step S304: Based on the K-value sample data and parameter thresholds obtained from steps S301-S303, determine whether a user is a suspected anomalous user. Specifically, if the K-value sample data meets the following conditions: standard deviation < 0.4 and median > 0.2; or standard deviation within 0.4 to 0.6, minimum observation > 0, and interquartile range < 0.6, then the user is determined to be a suspected anomalous user. By introducing the K-value as an indicator of anomalous user criterion, the model calculation results can be verified more objectively, improving the accuracy and robustness of identification.

[0077] It should be noted that the above thresholds can be adjusted based on actual user electricity consumption. Setting reasonable parameter thresholds, such as standard deviation, median, minimum observation value, and interquartile range, can further refine the screening criteria for suspected abnormal users, ensuring the accuracy and reliability of identification. Users whose daily electricity consumption is consistently less than 1 unit should be included in continuous monitoring.

[0078] Step S305: Output the list of abnormal users. The list of abnormal users provides direct decision support for the power company. Based on this list, the power company can quickly locate users who may have abnormal electricity consumption behavior and then take corresponding measures.

[0079] Step S4: Perform data validation on the case set. By performing data validation on the case set, the performance of the model in real-world applications can be verified, which helps to identify potential problems and shortcomings in the model, and then to make targeted optimizations and improvements.

[0080] Specifically, by comparing the model's identification results with actual conditions, the accuracy of the identification can be evaluated, and the model's parameters and thresholds can be adjusted accordingly to improve accuracy. By testing under different scenarios and conditions, the model's ability to identify various abnormal electricity consumption behaviors can be assessed, and the model can be optimized to better adapt to different electricity consumption environments and scenarios.

[0081] In summary, the robust outlier-based abnormal electricity consumption detection result verification method provided in this embodiment identifies outliers in user electricity consumption behavior within low-voltage distribution areas. By deeply extracting user electricity consumption behavior characteristics, it uncovers abnormal electricity consumption behaviors within users. Furthermore, by optimizing the algorithm to improve robustness, it ensures the stability and reliability of detection results even in the face of various complex situations such as data uncertainty, noise, and external interference. In addition, by incorporating the impact of actual user electricity consumption changes on distribution area line loss changes, an indicator influence degree K is constructed as an abnormal user criterion. This objectively verifies the model calculation results and comprehensively analyzes user status, improving identification efficiency, accuracy, and robustness.

[0082] Please refer to Figure 4 In one embodiment, the present invention provides a robust outlier anomaly power consumption detection result verification system, comprising:

[0083] The data acquisition and preprocessing module 100 is used to acquire electricity consumption data of users in the distribution area based on the metering automation system, and to perform data cleaning and preprocessing.

[0084] The feature selection module 200 is used to analyze the feature indicators related to the user's daily electricity consumption and the line loss of the transformer area using screening and embedding methods, and to perform heuristic search and factor analysis between features to select key features.

[0085] The model building and output module 300 is used to build an outlier algorithm model based on key features, train the model to find suitable parameter thresholds, and output suspected users with abnormal electricity consumption behavior.

[0086] The data validation module 400 is used to perform data validation on the case set.

[0087] The robust outlier abnormal power consumption detection result verification system provided in this application embodiment can execute the above method embodiment, and its implementation principle and technical effect are similar, so it will not be described again here.

[0088] Those skilled in the art will understand that various aspects of the present invention can be implemented as systems, methods, or program products. Therefore, various aspects of the present invention can be specifically implemented in the following forms: entirely hardware implementations, entirely software implementations (including firmware, microcode, etc.), or implementations combining hardware and software aspects, collectively referred to herein as “circuits,” “modules,” or “systems.”

[0089] Any one or more of the modules, submodules, units, and subunits according to this embodiment, or at least part of the functions of any one or more of them, can be implemented in one module. Any one or more of the modules, submodules, units, and subunits according to this embodiment can be implemented by dividing them into multiple modules. Any one or more of the modules, submodules, units, and subunits according to this embodiment can be at least partially implemented as hardware circuits, such as field-programmable gate arrays (FPGAs), programmable logic arrays (PLAs), systems-on-a-chip, systems-on-a-substrate, systems-on-package, application-specific integrated circuits (ASICs), or implemented by hardware or firmware in any other reasonable manner by integrating or packaging circuits, or implemented in software, hardware, and firmware, or in any appropriate combination of any of these three implementation methods. Alternatively, one or more of the modules, submodules, units, and subunits according to this embodiment can be at least partially implemented as computer program modules, which, when run, can perform corresponding functions.

[0090] In one embodiment, the present invention provides a robust outlier anomaly power consumption detection result verification device, which may include at least one processing unit and at least one storage unit. The storage unit stores program code, which, when executed by the processing unit, causes the processing unit to perform the steps described above in the robust outlier anomaly power consumption detection result verification according to various exemplary embodiments of the present invention. For example, the processing unit may perform, as follows: Figure 1 The flowchart illustrates the robust outlier anomaly power consumption detection result verification method.

[0091] In one embodiment, the present invention provides a computer-readable medium storing computer-executable instructions for performing the steps in the robust outlier anomaly power consumption detection result verification method according to various exemplary embodiments of the present invention described above.

[0092] A readable medium can be a readable signal medium or a readable storage medium. A readable storage medium can be, for example—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. A readable signal medium can include data signals propagated in baseband or as part of a carrier wave, carrying readable program code. A readable signal medium can also be any readable medium other than a readable storage medium, capable of transmitting, propagating, or transmitting a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the readable medium can be transmitted using any suitable medium, including—but not limited to—wireless, wired, optical fiber, RF, etc., or any suitable combination thereof.

[0093] Furthermore, although the operations of the method of the present invention are described in a specific order in the accompanying drawings, this does not require or imply that these operations must be performed in that specific order, or that all the operations shown must be performed to achieve the desired result. Additionally or alternatively, certain steps may be omitted, multiple steps may be combined into one step, and / or one step may be broken down into multiple steps.

[0094] It should be understood that those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims.

[0095] The present invention has been described above with reference to the accompanying drawings. Obviously, the implementation of the present invention is not limited to the above-described manner. Any improvements made using the inventive concept and technical solution of the present invention, or the direct application of the inventive concept and technical solution of the present invention to other situations without modification, are all within the protection scope of the present invention.

Claims

1. A robust-based outlier anomaly electricity consumption detection result verification method, characterized in that, The method comprises the following steps: Step S1, obtaining the user power consumption data of a transformer area based on a metering automation system, and performing data cleaning and preprocessing; Step S2, using a screening method and an embedding method to analyze characteristic indexes related to the daily power consumption of a user and the line loss of a transformer area, and performing heuristic search and factor analysis between the characteristic indexes to select key characteristics; Step S3, constructing an outlier algorithm model based on the key characteristics, and performing model training to find suitable parameter thresholds and output suspicious users with abnormal power consumption behaviors; Step S4, performing data verification on a case set; Step S1 specifically comprises the following sub-steps: Step S101, removing daily power consumption data of a transformer area with a power supply amount ≤0, and removing daily power consumption data with a line loss rate <-2% or a line loss rate >40%; Step S102, combining a horizontal comparison and a vertical comparison of a transformer area to perform interpolation processing on missing daily power consumption data of a user; Step S103, using a confidence interval to identify and correct abnormal power consumption of a user; In step S102, when the number of consecutive missing days of daily power consumption of a user is ≤7 days, a cubic spline interpolation method is used to interpolate the missing daily power consumption data of the user, and the power supply amount of the transformer area after interpolation meets the condition of the line loss rate in step S101; The characteristic indexes include daily power consumption, daily power consumption change, line loss power, line loss power change rate, K value, and restored power, the K value is an index influence degree, and the restored power is actual power of a user; The K value is calculated based on the actual power of a user on the same day, the actual power of a user on the previous day, the line loss amount of a transformer area on the same day, and the line loss amount of a transformer area on the previous day; The formula for calculating the K value based on the actual power of a user on the same day, the actual power of a user on the previous day, the line loss amount of a transformer area on the same day, and the line loss amount of a transformer area on the previous day is as follows:

2. The robust-based outlier anomaly electricity usage detection result verification method of claim 1, wherein, In step S103, the formula of the confidence interval is as follows: ； Wherein, the mean is the average value of the user power consumption data, the standard deviation is the measure of the dispersion degree of the user power consumption data, z is the z-score of the standard normal distribution, and n is the number of data points.

3. The robust-based outlier anomaly electricity usage detection result verification method of claim 1, wherein, User power consumption data outside the confidence interval range is identified as abnormal power consumption data, and is corrected to the boundary value of the confidence interval, and the power supply amount of the transformer area is also corrected accordingly.

4. The robust-based outlier anomaly electricity usage detection result verification method of claim 1, wherein, The restored power is obtained by constructing a power conservation equation and using the least square method to solve the user error rate.

5. The robust-based outlier anomaly electricity usage detection result verification method of claim 1, wherein, Step S3 specifically comprises the following sub-steps: Step S301, selecting power consumption jump sample data; Step S302, selecting abnormal power consumption sample data; Step S303, removing sample data with abnormal line loss power; Step S304, determining whether a user is an abnormal suspicious user according to the K value sample data selected in steps S301-S303 and the parameter threshold; Step S305, outputting an abnormal user list.

6. A robust outlier anomaly electricity consumption detection result verification system that performs the robust outlier anomaly electricity consumption detection result verification method according to any one of claims 1-5, characterized in that, It comprises: A data acquisition and preprocessing module, configured to obtain user power consumption data of a transformer area based on a metering automation system, and perform data cleaning and preprocessing; The feature selection module is configured to analyze feature indexes related to the daily power consumption of the user and the line loss of the transformer area by using a screening method and an embedding method, perform heuristic search on the feature indexes and factor analysis between the features, and select key features; The model construction and output module is configured to construct an outlier algorithm model based on the key features, find suitable parameter thresholds by model training, and output suspicious users with abnormal power consumption behaviors; The data verification module is configured to perform data verification on the case set.

Images