Grassroots-oriented electricity stealing identification method and device
By analyzing electricity characteristics data from historical stealing user data and training the intelligent recognition model of electricity stealing users, the problem of insufficient accuracy of electricity recognition has been solved, and the accuracy of power theft recognition and lower misjudgment are achieved. Rate.
Patent Information
- Application Number
- CN202411288573.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-14
- Publication Date
- 2025-01-21
AI Technical Summary
The existing technology is difficult to effectively identify and prevent power theft, resulting in serious economic losses from power supply companies.
By using the intelligent analysis model of power stealing, the first -dimensional first user electricity characteristics data are analyzed from a large number of historical stealing user data, and these characteristic data are used to train the intelligent recognition model of electricity theft users to improve the accuracy of electricity theft recognition.
This method can significantly improve the accuracy of electricity theft identification, minimize misjudgments and omissions, and effectively protect the economic interests of power supply enterprises.
Smart Images
Figure CN119337266A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of electricity theft identification, and in particular to a grassroots-oriented electricity theft identification method and device. Background Art
[0002] With the rapid development of the power industry and the improvement of its intelligence level, the accuracy and security of power metering systems have become increasingly prominent. As a major problem in the field of power metering, electricity theft has brought serious economic losses to power supply companies.
[0003] Therefore, how to identify electricity stealing users has become a technical problem that needs to be solved urgently by those skilled in the art. Summary of the invention
[0004] The present application provides a method and device for identifying electricity theft at the grassroots level. In the method, an intelligent analysis model for electricity theft is used to analyze a large amount of historical electricity theft user data to obtain multi-dimensional electricity usage characteristic data of a first user, and the electricity usage characteristic data of the first user is used to train an intelligent identification model for electricity theft users. The intelligent identification model for electricity theft users can improve the accuracy of electricity theft identification and minimize misjudgment and missed judgments.
[0005] In a first aspect, a method for identifying electricity theft at the grassroots level is provided, comprising:
[0006] A large amount of historical electricity theft user data is analyzed using an electricity theft intelligent analysis model to obtain the first user's electricity consumption characteristic data;
[0007] Using the electricity consumption characteristic data of the first user, training an intelligent identification model for electricity theft users;
[0008] Determine the power consumption characteristic data of a second user of the current user;
[0009] The power consumption characteristic data of the second user of the current user is input into the intelligent identification model of the power stealing user to determine whether the user is stealing power.
[0010] In some embodiments, the electricity theft intelligent analysis model includes an anti-electricity theft intelligent recognition model, a high-price connection low-abnormal recognition model and an electricity consumption behavior feature intelligent analysis model.
[0011] In some embodiments, the first user's electricity consumption characteristic data corresponding to the anti-electricity theft intelligent identification model includes voltage and current abnormal information, secondary side shunt abnormal information, primary side shunt abnormal information, power factor abnormal information, no-load loss and reverse current power abnormal information, discrete coefficient abnormal information, power abnormal information, phase angle abnormal information and low-voltage user electricity theft information.
[0012] In some embodiments, the first user's electricity consumption characteristic data corresponding to the high-price connection low abnormality identification model includes a suspicion coefficient of an abnormal user.
[0013] In some embodiments, the first user's electricity consumption characteristic data corresponding to the intelligent analysis model of electricity consumption behavior characteristics includes similarity information with the load curve of the industry to which it belongs, user's own correlation information, similarity information with the current, standardized unit power, and electricity quantity of the preset user, and electricity consumption model information.
[0014] In some embodiments, the power consumption behavior feature intelligent analysis model includes a user power consumption pattern recognition model, and the user model information includes an abnormal power consumption pattern;
[0015] The step of analyzing a large amount of historical electricity theft user data using an electricity theft intelligent analysis model to obtain the first user's electricity consumption characteristic data includes:
[0016] A large amount of historical electricity theft user data is input into the user electricity usage pattern recognition model to output user model information; wherein, the user electricity usage pattern recognition model is used to preprocess and normalize the dedicated transformer user load data, and the normalized user load data is clustered using the K-means clustering algorithm, each cluster represents a user pattern information, and the electricity usage pattern of a single user in the past year is constructed through the clustering results; wherein, if the distance from the dedicated transformer user load data to the cluster center point is greater than the cluster threshold radius, the user pattern information is determined to be an abnormal electricity usage pattern.
[0017] In some embodiments, the step of using the first user's electricity usage characteristic data to train an intelligent identification model for electricity theft users includes:
[0018] Conducting exploration and analysis on the electricity consumption characteristic data of the first user to eliminate users who are unlikely to engage in electricity theft;
[0019] Preprocessing the first user's power consumption characteristic data after exploration and analysis to obtain preprocessed first user's power consumption characteristic data;
[0020] Splitting the preprocessed first user power consumption characteristic data into a test sample and a training sample according to a ratio;
[0021] The training samples are used to train the electricity theft user intelligent identification model of the RNN algorithm including the multi-step fusion mechanism, and the trained electricity theft user intelligent identification model is tested using the test samples. When the test result meets expectations, the current trained electricity theft user intelligent identification model is determined as the electricity theft user intelligent identification model.
[0022] In a second aspect, a grassroots electricity theft identification device is provided, comprising:
[0023] An analysis unit, used to analyze a large amount of historical electricity theft user data using an electricity theft intelligent analysis model to obtain electricity consumption characteristic data of a first user;
[0024] A training unit, configured to train an intelligent identification model for electricity theft users by using the electricity consumption characteristic data of the first user;
[0025] A determination unit, used to determine the power consumption characteristic data of a second user of the current user;
[0026] The input unit is used to input the power consumption characteristic data of the second user of the current user into the intelligent identification model of the power stealing user to determine whether the user steals power.
[0027] In a third aspect, an electronic device is provided, comprising: at least one processor; and a memory communicatively connected to the at least one processor;
[0028] The memory stores instructions that can be executed by the at least one processor, and the instructions are configured to execute the above method.
[0029] According to a fourth aspect, a computer-readable storage medium is provided, storing computer-executable instructions for executing the method.
[0030] In the above-mentioned embodiment, a method and device for identifying electricity theft at the grassroots level are provided. In this method, a large amount of historical electricity theft user data is analyzed using an intelligent analysis model for electricity theft to obtain multi-dimensional first user electricity usage characteristic data, and the first user electricity usage characteristic data is used to train an intelligent identification model for electricity theft users. The intelligent identification model for electricity theft users can improve the accuracy of electricity theft identification and minimize misjudgment and missed judgment. The method includes analyzing a large amount of historical electricity theft user data using an intelligent analysis model for electricity theft to obtain first user electricity usage characteristic data; using the first user electricity usage characteristic data to train an intelligent identification model for electricity theft users; determining the second user electricity usage characteristic data of the current user; and inputting the second user electricity usage characteristic data of the current user into the intelligent identification model for electricity theft users to determine whether the user has stolen electricity. BRIEF DESCRIPTION OF THE DRAWINGS
[0031] Figure 1 A flowchart of a grassroots electricity theft identification method is exemplarily shown according to some embodiments;
[0032] Figure 2 A schematic structural diagram of a grassroots electricity theft identification device provided according to some embodiments is exemplarily shown. DETAILED DESCRIPTION
[0033] In order to make the purpose and implementation method of the present application clearer, the exemplary implementation method of the present application will be clearly and completely described below in conjunction with the drawings in the exemplary embodiments of the present application. Obviously, the described exemplary embodiments are only part of the embodiments of the present application, rather than all the embodiments.
[0034] It should be noted that the brief description of terms in this application is only for the convenience of understanding the embodiments described below, and is not intended to limit the embodiments of this application. Unless otherwise specified, these terms should be understood according to their common and usual meanings.
[0035] The terms "first", "second", "third", etc. in the specification and claims of this application and the above drawings are used to distinguish similar or similar objects or entities, and do not necessarily mean to limit a specific order or sequence, unless otherwise noted. It should be understood that the terms used in this way can be interchangeable under appropriate circumstances.
[0036] The terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover but not exclude inclusion, for example, a product or device comprising a list of components is not necessarily limited to all the components expressly listed but may include other components not expressly listed or inherent to such product or device.
[0037] With the rapid development of the power industry and the improvement of its intelligence level, the accuracy and safety of power metering systems have become increasingly prominent. As a major problem in the field of power metering, electricity theft has brought serious economic losses to power supply companies. Therefore, how to identify electricity theft users has become a technical problem that technicians in this field need to solve urgently.
[0038] In order to solve the above technical problems, an embodiment of the present application provides a grassroots-oriented electricity theft identification method, in which an intelligent electricity theft analysis model is used to analyze a large amount of historical electricity theft user data to obtain multi-dimensional first user electricity usage characteristic data, and the first user's electricity usage characteristic data is used to train an intelligent electricity theft user identification model. The intelligent electricity theft user identification model can improve the accuracy of electricity theft identification and minimize false positives and missed positives.
[0039] Figure 1 The flowchart of a method for handling abnormal line loss according to some embodiments is exemplarily shown. The method includes S100-S400.
[0040] S100: Analyze a large amount of historical electricity theft user data using an electricity theft intelligent analysis model to obtain electricity consumption characteristic data of a first user.
[0041] In some embodiments, the electricity theft intelligent analysis model includes an anti-electricity theft intelligent recognition model, a high-price connection low-abnormal recognition model and an electricity consumption behavior feature intelligent analysis model.
[0042] The anti-electricity theft intelligent identification model is used to classify and organize various types of electricity consumption characteristics and electrical parameters based on data mining analysis through a large amount of historical electricity theft user data accumulated in history, and give corresponding judgment results. Then, many potential important factors, facts, associations and other valuable information contained in these data are extracted to analyze the correlation between electricity theft incidents and other factors (such as power, electricity, phase angle, etc.).
[0043] In some embodiments, the anti-electricity theft intelligent identification model includes a voltage and current anomaly identification model, a secondary side shunt anomaly identification model, a primary side shunt anomaly identification model, a power factor anomaly identification model, a no-load loss and reverse current power anomaly identification model, a discrete coefficient anomaly identification model, a power anomaly identification model, a phase angle anomaly identification and a low-voltage user electricity theft identification model.
[0044] The first user's electricity consumption characteristic data corresponding to the anti-electricity theft intelligent identification model includes voltage and current abnormal information, secondary side shunt abnormal information, primary side shunt abnormal information, power factor abnormal information, no-load loss and reverse current power abnormal information, discrete coefficient abnormal information, power abnormal information, phase angle abnormal information and low-voltage user electricity theft information.
[0045] Abnormal voltage and current information includes: three-phase three-wire voltage unbalance, three-phase four-wire voltage unbalance, low voltage, three-phase current unbalance, micro-current unbalance, and current loss.
[0046] The secondary side shunt abnormal information includes: A-phase CT secondary side shunt, B-phase CT secondary side shunt, C-phase CT secondary side shunt, and three-phase three-wire secondary side asymmetric shunt.
[0047] The primary side shunt abnormal information includes: A phase CT primary side shunt, B phase CT primary side shunt, C phase CT primary side shunt, and three-phase three-wire primary side asymmetric shunt.
[0048] Abnormal power factor information includes: low power factor.
[0049] The no-load loss and reverse current power abnormality information includes: daily power abnormality, no-load loss abnormality, reverse current or power.
[0050] The information of abnormal dispersion coefficient includes: abnormal power factor dispersion coefficient and abnormal power dispersion coefficient.
[0051] Abnormal power information includes: abnormal weight ratio of user power difference and abnormal user power ratio.
[0052] Phase angle abnormality information includes: collected phase angle abnormality, estimated phase angle abnormality, and three-phase active reverse polarity.
[0053] The information on electricity theft by low-voltage users includes: cover opening alarm records, electricity dispersion coefficient, electricity deviation rate, line loss data, electricity and line loss change difference, electricity and line loss change rate, supply and sales electricity change deviation rate, weight ratio of electricity change difference and line loss change difference, user electricity difference weight ratio, time series deviation rate dispersion coefficient, and current and voltage model pilot.
[0054] The "high price to low connection" in the high price to low connection anomaly identification model refers to the behavior of connecting a load with a higher electricity price to a load access point with a lower electricity price in the power grid. Analyzing this behavior can reduce the company's electricity bill losses.
[0055] The first user's electricity consumption characteristic data corresponding to the high-price connection low abnormality identification model includes a suspicion coefficient of an abnormal user.
[0056] In some embodiments, the high-price-to-low anomaly recognition model includes a residential electricity price anomaly recognition model and an agricultural production electricity price anomaly recognition model.
[0057] The suspicion coefficient of abnormal users includes: filtering out abnormal users according to abnormal residential electricity price rules and setting suspicion coefficients of abnormal users; and filtering out abnormal users according to abnormal agricultural production electricity price rules and setting suspicion coefficients of abnormal users.
[0058] In some embodiments, the power consumption behavior feature intelligent analysis model includes a power consumption trajectory anomaly analysis model, a user's own curve anomaly analysis model, a similarity anomaly model and a user power consumption pattern recognition model.
[0059] The first user's electricity consumption characteristic data corresponding to the intelligent analysis model of electricity consumption behavior characteristics includes similarity information with the load curve of the industry to which it belongs, user's own correlation information, similarity information with the current, standardized unit power and electricity of the preset user, and electricity consumption model information.
[0060] In some embodiments, the power consumption behavior feature intelligent analysis model includes a power consumption trajectory abnormality analysis model;
[0061] The step of analyzing a large amount of historical electricity theft user data using an electricity theft intelligent analysis model to obtain the first user's electricity consumption characteristic data includes:
[0062] A large amount of historical electricity theft user data is input into the electricity consumption trajectory anomaly analysis model to output similarity information with the load curve of the industry to which it belongs; wherein the electricity consumption trajectory anomaly analysis model is used to use a distance algorithm or a clustering algorithm to analyze the similarity between the user's electricity consumption trajectory and the industry load curve, and determine the similarity information with the load curve of the industry to which it belongs.
[0063] In some embodiments, the power consumption behavior feature intelligent analysis model further includes a user's own curve anomaly analysis model;
[0064] The step of analyzing a large amount of historical electricity theft user data using an electricity theft intelligent analysis model to obtain the first user's electricity consumption characteristic data also includes:
[0065] A large amount of historical electricity theft user data is used to determine the user's own correlation information by using the user's own curve anomaly analysis model; wherein the user's own curve anomaly analysis model is used to generate a user autocorrelation model and filter out the user's own correlation information obtained by comparing the user's current electricity consumption with the user's own historical electricity consumption on a year-on-year or month-on-month basis from the user's autocorrelation model; the user autocorrelation model is established through the linear dependence between the user's electricity consumption at time T and the electricity consumption at any interval K from the current time, according to the linear dependence, and is established by removing noise data in combination with the residual sequence method.
[0066] In the embodiment of the present application, regardless of the type of special transformer user, the user's power consumption or load curve always has autocorrelation, and the user's current data is provided with a year-on-year and month-on-month comparison with its own historical data, and the linear dependence (periodic interval) between the user's power consumption at time T and the power consumption at any interval K from the current time is examined. By establishing an autocorrelation coefficient and combining the residual sequence method to eliminate noise data, a user autocorrelation model is established.
[0067] In some embodiments, the power consumption behavior feature intelligent analysis model further includes a similarity anomaly model;
[0068] The step of analyzing a large amount of historical electricity theft user data using an electricity theft intelligent analysis model to obtain the first user's electricity consumption characteristic data also includes:
[0069] A large amount of historical electricity theft user data is input into the similarity anomaly model to output similarity information of current, standardized unit power and electricity quantity with the preset users; wherein the similarity anomaly model is used for similar user results classified by the clustering algorithm, and users matching the same region, same industry, same multiple, same capacity and similar current data are compared and detected with the abnormal users to be analyzed, and the similarity of current, standardized unit power and electricity quantity is judged to determine the similarity information of current, standardized unit power and electricity quantity with the preset users.
[0070] In some embodiments, the power consumption behavior feature intelligent analysis model further includes a user power consumption pattern recognition model, and the user model information includes an abnormal power consumption pattern;
[0071] The step of analyzing a large amount of historical electricity theft user data using an electricity theft intelligent analysis model to obtain the first user's electricity consumption characteristic data also includes:
[0072] A large amount of historical electricity theft user data is input into the user electricity usage pattern recognition model to output user model information; wherein, the user electricity usage pattern recognition model is used to preprocess and normalize the dedicated transformer user load data, and the normalized user load data is clustered using the K-means clustering algorithm, each cluster represents a user pattern information, and the electricity usage pattern of a single user in the past year is constructed through the clustering results; wherein, if the distance from the dedicated transformer user load data to the cluster center point is greater than the cluster threshold radius, the user pattern information is determined to be an abnormal electricity usage pattern.
[0073] S200: Using the first user's electricity usage characteristic data, train an intelligent identification model for electricity theft users.
[0074] In the embodiment of the present application, the intelligent analysis model for electricity theft includes multi-dimensional electricity consumption characteristic data of the first user formed by the data output by the anti-electricity theft intelligent identification model, the high-price connection low abnormal identification model and the electricity consumption behavior characteristic intelligent analysis model, and the intelligent identification model for electricity theft users is used to train the intelligent identification model for electricity theft users, so that the intelligent identification model for electricity theft users can identify electricity theft with higher accuracy.
[0075] In some embodiments, the step of using the first user's electricity usage characteristic data to train an intelligent identification model for electricity theft users includes:
[0076] An exploratory analysis is performed on the electricity consumption characteristic data of the first user to eliminate users who are unlikely to have engaged in electricity theft.
[0077] In the embodiment of the present application, a whitelist is set, that is, users who are unlikely to steal electricity. When users in the whitelist appear in the first user's electricity usage feature data, they can be removed to avoid affecting the recognition effect of the trained intelligent recognition model for electricity stealing users.
[0078] The first user power consumption characteristic data after exploration and analysis is preprocessed to obtain the preprocessed first user power consumption characteristic data. In the embodiment of the present application, the preprocessing may be data cleaning, missing value processing and data transformation.
[0079] The preprocessed first user power consumption characteristic data is split into test samples and training samples according to a ratio. For example, 20% of the preprocessed first user power consumption characteristic data can be used as test samples, and 80% of the preprocessed first user power consumption characteristic data can be used as training samples.
[0080] The training samples are used to train the electricity theft user intelligent identification model of the RNN algorithm including the multi-step fusion mechanism, and the trained electricity theft user intelligent identification model is tested using the test samples. When the test result meets expectations, the current trained electricity theft user intelligent identification model is determined as the electricity theft user intelligent identification model.
[0081] In some embodiments, the RNN algorithm of the multi-step fusion mechanism can fuse data from multiple time steps together to extract more comprehensive information. At each time step, the RNN outputs a state vector that contains information about the current time step and the previous time step. By fusing these state vectors, a comprehensive representation that incorporates information from multiple time steps can be obtained.
[0082] S300, determining power consumption characteristic data of a second user of the current user;
[0083] S400: Inputting the electricity usage characteristic data of a second user of the current user into the electricity stealing user intelligent identification model to determine whether the user steals electricity.
[0084] In some embodiments, the method further comprises:
[0085] Screen out users who meet the preset requirements, take the user as the current user, and subsequently determine the second user power consumption characteristic data of the current user. In the embodiment of the present application, due to the large number of users, users who are more likely to have power theft problems can be screened out for focused identification.
[0086] In some embodiments, users who meet preset requirements can be screened using the super-capacity anomaly recognition model.
[0087] The overcapacity anomaly recognition model sets thresholds for monitoring based on the customer's electricity usage patterns and industry electricity usage characteristics. For example, the preset requirement is that the user's monthly total electricity consumption, monthly peak electricity consumption, and monthly valley electricity consumption are greater than the corresponding thresholds for several consecutive months.
[0088] When the total monthly electricity, monthly peak electricity, and monthly valley electricity cannot be obtained, the following algorithm can be used for prediction:
[0089] The prediction formula for non-time-sharing users of dedicated transformers is: total monthly electricity consumption > transformer billing capacity × 18 hours × 30 days × 120% (this value is set according to actual conditions).
[0090] The prediction formula for dedicated transformer time-sharing users is as follows: monthly peak electricity consumption > transformer billing capacity × 6 hours × 30 days × 120%; monthly flat electricity consumption > transformer billing capacity × 10 hours × 30 days × 120%; monthly valley electricity consumption > transformer billing capacity × 8 hours × 30 days × 120%.
[0091] In this way, if the total monthly electricity consumption, monthly peak electricity consumption and monthly valley electricity consumption are affected by certain factors, it is still possible to predict the situation and ultimately select users who meet the preset requirements.
[0092] In the above embodiment, a method for identifying electricity theft at the grassroots level is provided, in which an intelligent analysis model for electricity theft is used to analyze a large amount of historical electricity theft user data to obtain multi-dimensional first user electricity usage characteristic data, and the first user electricity usage characteristic data is used to train an intelligent identification model for electricity theft users, and the intelligent identification model for electricity theft users can improve the accuracy of electricity theft identification and minimize misjudgment and missed judgment. The method includes analyzing a large amount of historical electricity theft user data using an intelligent analysis model for electricity theft to obtain first user electricity usage characteristic data; using the first user electricity usage characteristic data to train an intelligent identification model for electricity theft users; determining the second user electricity usage characteristic data of the current user; and inputting the second user electricity usage characteristic data of the current user into the intelligent identification model for electricity theft users to determine whether the user has stolen electricity.
[0093] Further, as Figure 1 The specific implementation of the method, the embodiment of the present application provides a grassroots electricity theft identification device, such as Figure 2 As shown, including:
[0094] An analysis unit 201 is used to analyze a large amount of historical electricity theft user data using an electricity theft intelligent analysis model to obtain electricity consumption characteristic data of a first user;
[0095] A training unit 202 is used to train an intelligent identification model for electricity theft users by using the electricity consumption characteristic data of the first user;
[0096] A determination unit 203, configured to determine the power consumption characteristic data of a second user of the current user;
[0097] The input unit 204 is used to input the power consumption characteristic data of the second user of the current user into the power stealing user intelligent identification model to determine whether the user steals power.
[0098] In a specific application scenario, the power consumption behavior feature intelligent analysis model includes a user power consumption pattern recognition model, and the user model information includes an abnormal power consumption pattern;
[0099] Analysis unit, including:
[0100] The first input unit is used to input a large amount of historical electricity theft user data into the user electricity usage pattern recognition model to output user model information; wherein, the user electricity usage pattern recognition model is used to preprocess and normalize the dedicated transformer user load data, and use the K-means clustering algorithm to cluster the normalized user load data, each cluster represents a user pattern information, and the electricity usage pattern of a single user in the past year is constructed through the clustering results; wherein, if the distance from the dedicated transformer user load data to the cluster center point is greater than the threshold radius of the cluster, the user pattern information is determined to be an abnormal electricity usage pattern.
[0101] In a specific application scenario, the training unit includes:
[0102] A removal unit, configured to perform exploration and analysis on the power consumption characteristic data of the first user, so as to remove users who are unlikely to engage in power theft;
[0103] A preprocessing unit, configured to preprocess the first user's power consumption characteristic data after exploration and analysis to obtain the preprocessed first user's power consumption characteristic data;
[0104] A splitting unit, used for splitting the preprocessed first user power consumption characteristic data into a test sample and a training sample according to a ratio;
[0105] A generating unit is used to train the electricity thief user intelligent identification model of the RNN algorithm including the multi-step fusion mechanism using the training samples, and to test the trained electricity thief user intelligent identification model using the test samples. When the test result meets the expectation, the current trained electricity thief user intelligent identification model is determined as the electricity thief user intelligent identification model.
[0106] It should be noted that for other corresponding descriptions of the functional units involved in the grassroots electricity theft identification device provided in this embodiment, please refer to Figure 1 The corresponding description in will not be repeated here.
[0107] Based on the above Figure 1 The method shown in the embodiment of the present application accordingly provides a storage medium on which a computer program is stored, and when the program is executed by a processor, the above-mentioned Figure 1 The utility model shows a grassroots-oriented electricity theft identification method.
[0108] Based on this understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), and includes a number of instructions for enabling an electronic device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each implementation scenario of the present application.
[0109] Based on the above Figure 1 The method shown, and Figure 2 In order to achieve the above-mentioned purpose, the embodiment of the present application also provides a physical device for grassroots electricity theft identification, which can be a computer, a smart phone, a tablet computer, a smart watch, a server, or a network device, etc. The physical device includes a storage medium and a processor; the storage medium is used to store a computer program; the processor is used to execute the computer program to achieve the above-mentioned Figure 1 The line loss abnormality handling method is shown.
[0110] Optionally, the physical device may also include a user interface, a network interface, a camera, a radio frequency (RF) circuit, a sensor, an audio circuit, a WI-FI module, etc. The user interface may include a display, an input unit such as a keyboard, etc., and the optional user interface may also include a USB interface, a card reader interface, etc. The network interface may optionally include a standard wired interface, a wireless interface (such as a WI-FI interface), etc.
[0111] Those skilled in the art will appreciate that the electronic device structure provided in this embodiment does not limit the electronic device, and may include more or fewer components, or a combination of certain components, or different component arrangements.
[0112] The storage medium may also include an operating device and a network communication module. The operating device is a program that manages and saves the hardware and software resources of the electronic device, and supports the operation of the information processing program and other software and / or programs. The network communication module is used to realize the communication between the various controls inside the storage medium, and the communication with other hardware and software in the physical device.
[0113] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present application can be implemented by means of software plus a necessary general hardware platform, or by hardware.
[0114] Those skilled in the art will appreciate that the accompanying drawings are only schematic diagrams of a preferred implementation scenario, and the units or processes in the accompanying drawings are not necessarily necessary for implementing the present application. Those skilled in the art will appreciate that the units in the devices in the implementation scenario can be distributed in the devices of the implementation scenario according to the description of the implementation scenario, or can be changed accordingly and located in one or more devices different from the present implementation scenario. The units of the above-mentioned implementation scenarios can be combined into one unit, or can be further split into multiple sub-units.
[0115] The above serial numbers of this application are only for description and do not represent the advantages and disadvantages of the implementation scenarios. The above disclosure is only a few specific implementation scenarios of this application, but this application is not limited to them, and any changes that can be thought of by technicians in this field should fall within the scope of protection of this application.
Claims
1. A method for identifying electricity theft at the grassroots level, characterized in that: include: A large amount of historical electricity theft user data is analyzed using an electricity theft intelligent analysis model to obtain the first user's electricity consumption characteristic data; Using the electricity consumption characteristic data of the first user, training an intelligent identification model for electricity theft users; Determine the power consumption characteristic data of a second user of the current user; The power consumption characteristic data of the second user of the current user is input into the intelligent identification model of the power stealing user to determine whether the user is stealing power.
2. The method according to claim 1, characterized in that The electricity theft intelligent analysis model includes an anti-electricity theft intelligent recognition model, a high-price connection low-abnormal recognition model and an electricity consumption behavior feature intelligent analysis model.
3. The method according to claim 2, characterized in that The first user's electricity consumption characteristic data corresponding to the anti-electricity theft intelligent identification model includes voltage and current abnormal information, secondary side shunt abnormal information, primary side shunt abnormal information, power factor abnormal information, no-load loss and reverse current power abnormal information, discrete coefficient abnormal information, power abnormal information, phase angle abnormal information and low-voltage user electricity theft information.
4. The method according to claim 2, characterized in that: The first user's electricity consumption characteristic data corresponding to the high-price connection low abnormality identification model includes a suspicion coefficient of an abnormal user.
5. The method according to claim 2, characterized in that: The first user's electricity consumption characteristic data corresponding to the intelligent analysis model of electricity consumption behavior characteristics includes similarity information with the load curve of the industry to which it belongs, user's own correlation information, similarity information with the current, standardized unit power and electricity of the preset user, and electricity consumption model information.
6. The method according to claim 2, characterized in that The power consumption behavior feature intelligent analysis model includes a user power consumption pattern recognition model, and the user model information includes an abnormal power consumption pattern; The step of analyzing a large amount of historical electricity theft user data using an electricity theft intelligent analysis model to obtain the first user's electricity consumption characteristic data includes: A large amount of historical electricity theft user data is input into the user electricity usage pattern recognition model to output user model information; wherein, the user electricity usage pattern recognition model is used to preprocess and normalize the dedicated transformer user load data, and the normalized user load data is clustered using the K-means clustering algorithm, each cluster represents a user pattern information, and the electricity usage pattern of a single user in the past year is constructed through the clustering results; wherein, if the distance from the dedicated transformer user load data to the cluster center point is greater than the cluster threshold radius, the user pattern information is determined to be an abnormal electricity usage pattern.
7. The method according to claim 2, characterized in that The step of using the first user's electricity consumption characteristic data to train an intelligent identification model for electricity theft users comprises: Conducting exploration and analysis on the electricity consumption characteristic data of the first user to eliminate users who are unlikely to engage in electricity theft; Preprocessing the first user's power consumption characteristic data after exploration and analysis to obtain preprocessed first user's power consumption characteristic data; Splitting the preprocessed first user power consumption characteristic data into a test sample and a training sample according to a ratio; The training samples are used to train the electricity theft user intelligent identification model of the RNN algorithm including the multi-step fusion mechanism, and the trained electricity theft user intelligent identification model is tested using the test samples. When the test result meets expectations, the current trained electricity theft user intelligent identification model is determined as the electricity theft user intelligent identification model.
8. A grassroots electricity theft identification device, characterized in that: include: An analysis unit, used to analyze a large amount of historical electricity theft user data using an electricity theft intelligent analysis model to obtain electricity consumption characteristic data of a first user; A training unit, configured to train an intelligent identification model for electricity theft users by using the electricity consumption characteristic data of the first user; A determination unit, used to determine the power consumption characteristic data of a second user of the current user; The input unit is used to input the power consumption characteristic data of the second user of the current user into the intelligent identification model of the power stealing user to determine whether the user steals power.
9. An electronic device, characterized in that: include: at least one processor; and, a memory communicatively coupled to the at least one processor; The memory stores instructions executable by the at least one processor, wherein the instructions are configured to execute the method according to any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that: Computer executable instructions are stored, and the computer executable instructions are used to execute the method according to any one of claims 1 to 7.