Electric energy meter current metering difference identification method and system

Abstract

The invention relates to the technical field of current difference identification, in particular to an electric energy meter current metering difference identification method and system. The method comprises the following steps of current acquisition network construction, multi-dimensional anomaly analysis, user behavior influence evaluation and model training. The accuracy and reliability of electric energy meter current metering difference identification are improved. Performing abnormal electric energy meter screening on the current acquisition network according to the electric energy meter current metering difference data to obtain abnormal electric energy meter screening data; performing current transformer performance detection on the abnormal electric energy meter screening data to generate abnormal electric energy meter hardware identification data; and performing electric energy meter firmware version verification on the abnormal electric energy meter screening data so as to generate abnormal electric energy meter software identification data. According to the method, the accuracy and reliability of current metering difference identification of the electric energy meter are improved through current acquisition network construction, multi-dimensional anomaly analysis, user behavior influence evaluation and model training.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of current difference identification, and in particular to a power meter current metering difference identification method and system. BACKGROUND

[0002] With the continuous development of smart grid and energy management technology, the accuracy and stability of power meters, which are the core devices for power metering and monitoring, play an important role in the power system. Traditional power meters are mainly mechanical and electronic, and usually use simple current and voltage sampling technology for metering. However, due to factors such as environmental temperature, equipment aging and power grid fluctuations, current metering may have errors, which in turn affects the accuracy of power settlement. In recent years, electronic power meters based on digital signal processing (DSP) and advanced sensor technology have rapidly popularized, achieving higher metering accuracy through high-precision sampling and complex metering algorithms. With the complexity of power load and the increase of distributed energy access, the nonlinearity and harmonic components of current waveform have increased significantly, making current metering difference identification an important research topic. In order to cope with these challenges, artificial intelligence, deep learning and other technologies have been gradually introduced by academia and industry to analyze real-time data and historical records of power meters, and to accurately identify and predict metering differences. However, the current collection of power meter current by traditional methods is mostly single-point sampling or simple distributed sampling, which cannot form a comprehensive collection network, and the existing technology usually only analyzes a single dimension (such as hardware failure or software problem), which makes it difficult to accurately locate the root cause of the metering difference, thereby reducing the accuracy and reliability of power meter current metering difference identification. SUMMARY

[0003] Therefore, it is necessary to provide a power meter current metering difference identification method and system to solve at least one of the above technical problems.

[0004] To achieve the above purpose, a power meter current metering difference identification method, the method comprising the following steps:

[0005] Step S1: Obtain power meter location information data; based on the power meter location information data, construct a current collection network for the power meter, generate a current collection network; collect power meter current data for the current collection network to obtain standard power meter current collection data; and identify power meter current metering differences for the standard power meter current collection data to generate power meter current metering difference data;

[0006] Step S2: Based on the current measurement difference data of the electricity meters, the current acquisition network is screened for abnormal electricity meters to obtain abnormal electricity meter screening data; the performance of the current transformer is tested on the abnormal electricity meter screening data to generate abnormal electricity meter hardware identification data; the electricity meter firmware version is verified on the abnormal electricity meter screening data to generate abnormal electricity meter software identification data; electromagnetic interference analysis is performed on the abnormal electricity meter screening data to generate abnormal electricity meter electromagnetic interference identification data; the abnormal electricity meter hardware identification data, abnormal electricity meter software identification data, and abnormal electricity meter electromagnetic interference identification data are integrated to generate electricity meter equipment performance difference identification data.

[0007] Step S3: Perform user electricity load fluctuation analysis on the abnormal electricity meter screening data to generate electricity load fluctuation curves; perform illegal electricity use detection on the electricity load fluctuation curves to generate illegal electricity use detection data and normal electricity use detection data; perform hierarchical electricity use behavior impact analysis on the illegal electricity use detection data and normal electricity use detection data to generate electricity meter electricity use behavior difference identification data.

[0008] Step S4: Train the model using the data on differences in the performance of the electricity meter equipment and the data on differences in the electricity consumption behavior of the electricity meter to generate a model for identifying differences in the current measurement of the electricity meter; import the data on differences in the performance of the electricity meter equipment and the data on differences in the electricity consumption behavior of the electricity meter into the model for identifying and classifying differences in the current measurement of the electricity meter, and generate a classification and matching report for the difference in the current measurement of the electricity meter to perform automated operation of identifying differences in the current measurement of the electricity meter.

[0009] This invention constructs a location-based current acquisition network to ensure comprehensive current acquisition and network coverage accuracy. The standardized current data collected improves the accuracy of metering discrepancy identification and reduces the impact of data inconsistencies and environmental interference on discrepancy recognition. By screening abnormal energy meters, problematic devices are quickly identified. Combining hardware performance testing, firmware version verification, and electromagnetic interference analysis, the causes of anomalies are analyzed from multiple dimensions, providing a reliable basis for device performance optimization. The integrated device performance difference data helps accurately identify metering errors caused by hardware or software. Abnormal electricity usage patterns are identified through electricity load fluctuation curve analysis. Illegal electricity usage detection provides technical support for preventing electricity theft or irregular electricity use. Layered electricity usage behavior impact analysis refines the impact of user behavior on metering discrepancies, laying the foundation for comprehensive problem identification. Combining device performance and user behavior difference data, model training improves the accuracy and reliability of metering discrepancy identification. Generating discrepancy identification reports through classification and matching automates and simplifies the identification process, reducing the cost and error of manual intervention. Therefore, this invention improves the accuracy and reliability of energy meter current metering discrepancy identification by constructing a current acquisition network, performing multi-dimensional anomaly analysis, assessing the impact of user behavior, and training a model.

[0010] Preferably, step S1 includes the following steps:

[0011] Step S11: Obtain the location information data of the electricity meter;

[0012] Step S12: Construct a current acquisition network for the electricity meter based on the location information data of the electricity meter to generate a current acquisition network; collect electricity meter current data from the current acquisition network at the same time interval to obtain electricity meter current acquisition data.

[0013] Step S13: Perform data preprocessing on the current acquisition data of the electricity meter to generate standard current acquisition data of the electricity meter. The data preprocessing includes data cleaning, data denoising, data missing value filling and standardization.

[0014] Step S14: Analyze the current measurement differences of the standard electricity meter data, generate current measurement difference data and current measurement non-difference data, and remove the current measurement non-difference data.

[0015] This invention achieves comprehensive coverage of electricity meter current data by acquiring meter location information and constructing a current acquisition network. Data acquisition at identical time intervals ensures temporal consistency, making it suitable for subsequent analysis. Data cleaning, noise reduction, missing value imputation, and standardization improve the accuracy and consistency of current acquisition data, providing high-quality input data for subsequent discriminant analysis. Metering discrepancies are identified by distinguishing between differential and non-differential data from standard current acquisition data, thus identifying metering anomalies. By eliminating non-differential data and focusing on analyzing differential data, the efficiency and relevance of data processing are optimized. The generation of metering discrepancy data provides crucial evidence for subsequent electricity meter performance evaluation and anomaly analysis, helping to promptly identify and resolve metering problems and ensure system stability and accuracy. The steps are closely interconnected and hierarchically structured, forming a logically clear and highly operable process that is easily integrated with existing power grid monitoring systems, improving overall system operating efficiency.

[0016] Preferably, the construction of a current acquisition network for electricity meters based on electricity meter location information data includes the following steps:

[0017] Geospatial mapping is performed on the location information data of electricity meters to generate spatial mapping data of electricity meter locations; based on the spatial mapping data of electricity meter locations, the current collection areas of electricity meters are divided to generate current collection area data of electricity meters.

[0018] Based on the data of the current collection area of ​​the electricity meter, the network topology is selected to obtain the network topology data of the electricity meter current; based on the network topology data of the electricity meter current, the regional electricity meter density is analyzed to generate regional electricity meter density data.

[0019] Based on the regional electricity meter density data, gateways are deployed for the electricity meter current acquisition area data to generate electricity meter acquisition gateway deployment data; the regional electricity meter density data is collected and transmitted centrally through the electricity meter acquisition gateway deployment data to generate a current acquisition network; electricity meter current data is collected from the current acquisition network at the same time intervals to obtain electricity meter current acquisition data.

[0020] This invention uses geospatial mapping of electricity meter location information to generate spatial mapping data of electricity meter locations. This data intuitively reflects the actual distribution of electricity meters and provides a scientific basis for subsequent regional division. Based on the spatial mapping data of electricity meter locations, the data acquisition area is divided to ensure reasonable division, reduce redundancy or gaps within areas, and improve the coverage and efficiency of current data acquisition. By selecting appropriate network topology structures (such as star, tree, or grid) based on regional data, electricity meter current network topology data is generated, which helps optimize network communication paths and reduce latency and data transmission costs. The generation of regional electricity meter density data provides a basis for the deployment of acquisition gateways, allowing gateways to be located closer to high-density areas, ensuring acquisition efficiency while reducing the number and cost of gateway devices. Centralized transmission of current acquisition data through gateways reduces data redundancy, lowers the complexity and operating cost of the acquisition system, and improves the stability and accuracy of data transmission. Data acquisition from the current acquisition network at the same time intervals ensures data time synchronization, which is helpful for subsequent time-series analysis and anomaly detection. The constructed current acquisition network enables efficient transmission of electricity meter current data over a large area, providing support for real-time monitoring and dynamic adjustment, and improving the flexibility and responsiveness of power grid management.

[0021] Preferably, step S14 includes the following steps:

[0022] The average current of the standard energy meter is calculated by analyzing the current data collected from the standard energy meter. The average current of the energy meter is then used as the reference current value, and the current measurement difference of the standard energy meter current data is calculated based on the reference current value to obtain the current measurement difference value.

[0023] The current metering difference value is judged according to the preset current metering difference threshold. When the current metering difference value is greater than or equal to the preset current metering difference threshold, the corresponding current metering difference value is marked as the current metering difference data of the energy meter.

[0024] When the current metering difference value is less than the preset current metering difference threshold, the corresponding current metering difference value is marked as non-difference data of the current meter, and the non-difference data of the current meter are removed from the data in the same time period.

[0025] This invention calculates the average current value from standard electricity meter current collection data to generate a representative average current value as a benchmark, ensuring the accuracy and scientific validity of current metering discrepancies calculation. Based on the benchmark current value and the current metering discrepancy calculation, it can quickly identify metering anomalies and generate discrepancy data, providing a reliable basis for further analysis of electricity meter performance. By comparing with a preset current metering discrepancy threshold, the metering discrepancy values ​​are divided into "discrepancy data" and "non-discrepancy data." This dynamic discrimination mechanism can adapt to different power grid areas and usage scenarios, improving the adaptability and flexibility of the method. The removal of non-discrepancy data from electricity meter current meters within the same time period reduces data storage and processing redundancy, providing a clearer data view for subsequent anomaly analysis and optimization, and reducing system computational complexity. The marking of discrepancy data and the removal of non-discrepancy data ensure more accurate location of metering problems, helping to quickly identify problematic electricity meters and their corresponding time periods, shortening problem diagnosis and maintenance time. The generated electricity meter current metering discrepancy data provides a key basis for the maintenance, repair, and optimization of abnormal electricity meters and metering systems, effectively improving the reliability of electricity metering and the efficiency of power grid operation. The data clarification and problem-focusing processing flow not only reduces the impact of data noise on the system, but also improves the accuracy and efficiency of metering anomaly monitoring, ensuring the efficient operation of the power grid management system.

[0026] Preferably, step S2 includes the following steps:

[0027] Step S21: Screen abnormal energy meters in the current acquisition network based on the difference data of current meter readings, and obtain abnormal energy meter screening data.

[0028] Step S22: Perform current transformer performance testing on the abnormal energy meter screening data to generate energy meter current transformer performance data; based on the energy meter current transformer performance data, perform energy meter internal circuit testing on the abnormal energy meter screening data to generate abnormal energy meter hardware identification data.

[0029] Step S23: Check the calibration parameters of the abnormal energy meter screening data to obtain the abnormal energy meter parameter calibration record; verify the energy meter firmware version of the abnormal energy meter screening data through the abnormal energy meter parameter calibration record to generate abnormal energy meter software identification data.

[0030] Step S24: Monitor the working environment of the abnormal energy meter screening data to generate abnormal energy meter environmental monitoring data; perform electromagnetic interference analysis on the abnormal energy meter environmental monitoring data to generate abnormal energy meter electromagnetic interference identification data; integrate the abnormal energy meter hardware identification data, abnormal energy meter software identification data, and abnormal energy meter electromagnetic interference identification data to generate energy meter equipment performance difference identification data.

[0031] This invention combines current acquisition networks with current metering difference data to accurately screen abnormal energy meters, generating abnormal energy meter screening data and providing a reliable target range for subsequent in-depth testing and analysis. Testing the current transformer performance and internal circuits of abnormal energy meters comprehensively assesses the operating status of the meter hardware, accurately pinpoints the cause of hardware failures or performance degradation, and generates hardware identification data to provide data support for maintenance. Calibration parameter checks and firmware version verification ensure the correctness of meter parameter settings and firmware versions, avoiding metering anomalies caused by software issues. The generation of software identification data provides a more comprehensive perspective for anomaly analysis. Operating environment monitoring and electromagnetic interference analysis can identify the potential impact of environmental factors on energy meter operation, such as excessively high temperature, humidity, or electromagnetic interference, generating electromagnetic interference identification data to help optimize the energy meter deployment environment. By integrating hardware identification data, software identification data, and electromagnetic interference identification data, performance difference identification data for energy meter equipment is generated, providing comprehensive and systematic diagnostic results for performance problems of abnormal energy meters. By integrating hardware, software, and environmental monitoring data, the causes of faulty energy meters can be categorized, such as hardware failures, software incompatibility, or environmental interference, providing a basis for prioritizing solutions. The precise screening and comprehensive analysis process reduces unnecessary repairs, improves the efficiency of faulty energy meter detection and repair, and contributes to enhancing the overall stability and reliability of the power grid. Performance difference identification data helps managers discover systemic problems, such as firmware defects or insufficient environmental adaptability in specific energy meter models, thereby supporting the optimization of energy meter products and the power grid system.

[0032] Preferably, step S22 includes the following steps:

[0033] Step S221: Perform linear feature analysis on the abnormal energy meter screening data to generate current transformer characteristic parameter data; perform spectral decomposition on the current transformer characteristic parameter data to generate current harmonic component data.

[0034] Step S222: Perform dynamic response modeling on the current harmonic component data to generate the current transformer performance data of the energy meter; perform fault tree analysis on the current transformer performance data of the energy meter to generate circuit node status data.

[0035] Step S223: Perform circuit simulation on the circuit node state data to generate circuit operation characteristic data; perform pattern recognition on the abnormal energy meter screening data based on the circuit operation characteristic data to generate abnormal energy meter hardware identification data.

[0036] This invention analyzes the linear characteristics of abnormal energy meters to extract key current transformer characteristic parameters and performs spectral decomposition to generate current harmonic component data. This captures potential nonlinear characteristics and performance anomalies, effectively identifying the underlying characteristics causing the problem. A dynamic response model is constructed based on the current harmonic component data, generating accurate current transformer performance data, which helps to comprehensively evaluate the dynamic operating capability of the energy meter's current transformer. The dynamic model reflects the equipment's behavior under different operating conditions, providing a reliable basis for predicting potential faults. Using fault tree analysis, the current transformer performance data is analyzed in depth to generate circuit node state data, helping to quickly locate fault points and establish causal chains of performance anomalies. This method supports rapid identification of root causes in complex circuit systems and optimizes maintenance processes. Through circuit simulation, the circuit node state data is simulated and analyzed to generate circuit operating characteristic data, accurately reproducing the circuit behavior of the energy meter in actual operation. The simulation results can verify the hardware's operational stability, discover potential fault hazards, and thus improve detection accuracy. Based on circuit operation characteristic data, pattern recognition methods are used to generate hardware identification data for abnormal energy meters, ensuring comprehensive identification and classification of hardware faults. This data-driven identification approach can capture complex patterns of hardware faults, enhancing fault prediction capabilities. Combining feature analysis, dynamic modeling, simulation, and pattern recognition, a complete performance testing and fault diagnosis process is formed. This process can quickly screen for abnormal energy meters with hardware problems, reduce invalid detections, and optimize the allocation of operation and maintenance resources.

[0037] Preferably, electromagnetic interference analysis of abnormal electricity meter environmental monitoring data includes:

[0038] Three-dimensional electromagnetic field scanning and mapping are performed on the environmental monitoring data of abnormal energy meters to generate electromagnetic field intensity distribution data; wavelet transform time-frequency decomposition is performed on the electromagnetic field intensity distribution data to generate electromagnetic interference spectrum data.

[0039] Multi-point triangulation is performed on the electromagnetic interference spectrum data to locate the electromagnetic interference source; ray tracing attenuation analysis is performed on the electromagnetic interference source location data to generate electromagnetic interference propagation characteristic data.

[0040] Electromagnetic field-circuit co-simulation was performed on electromagnetic interference propagation characteristic data to generate electromagnetic interference coupling mode data.

[0041] The electromagnetic interference coupling mode data is processed by neural network classification and recognition to generate electromagnetic interference identification data for abnormal energy meters.

[0042] This invention generates electromagnetic field intensity distribution data through three-dimensional electromagnetic field scanning and mapping, which comprehensively reflects the electromagnetic field distribution around the energy meter, helping to identify concentrated areas of electromagnetic interference and providing high-precision basic data for subsequent analysis. Accurate understanding of the electromagnetic environment helps to quickly locate abnormal areas and high-intensity interference sources. Wavelet transform time-frequency decomposition extracts electromagnetic interference spectrum data, clearly displaying the temporal and frequency characteristics of the interference. It can identify transient and periodic interference, accurately characterizing the characteristics of interference sources and providing a basis for classification and source tracing. Through multi-point triangulation positioning, the generated electromagnetic interference source location data can quickly determine the spatial location of the interference source, greatly shortening the problem investigation time. The precise positioning function supports rapid isolation of interference sources in complex electromagnetic environments, improving the efficiency of environmental monitoring. Ray tracing attenuation analysis generates electromagnetic interference propagation characteristic data that reveals the propagation path, attenuation law, and impact on the energy meter of the interference signal in space, providing a quantitative basis for assessing the interference intensity and range, and formulating effective shielding and isolation measures. Electromagnetic interference coupling mode data generated by electromagnetic field-circuit co-simulation reveals how the interference signal affects the internal circuitry of the energy meter, helping to deeply analyze the actual impact of electromagnetic interference. This simulation can simulate complex electromagnetic environments, providing a scientific basis for developing anti-interference design optimizations.

[0043] Preferably, step S3 includes the following steps:

[0044] Step S31: Collect user electricity consumption data from the abnormal energy meter screening data to obtain abnormal energy meter current data; perform load fluctuation analysis on the abnormal energy meter current data to generate electricity load fluctuation data; perform curve conversion on the electricity load fluctuation data to generate electricity load fluctuation curve.

[0045] Step S32: Calculate the slope of the power load fluctuation curve to obtain load fluctuation characteristic data; use the load fluctuation characteristic data to detect illegal electricity use of abnormal power meter current data, and generate illegal electricity use detection data and normal electricity use detection data.

[0046] Step S33: Based on the illegal electricity use detection data, conduct an impact assessment on the abnormal electricity meter current data to generate illegal electricity use behavior data; conduct peak and valley period electricity use behavior analysis on the normal electricity use detection data to generate normal electricity use behavior data.

[0047] Step S34: Integrate the illegal electricity use behavior data and the normal electricity use behavior data to generate electricity meter electricity use behavior difference identification data.

[0048] This invention combines user electricity consumption data collection with load fluctuation analysis and curve transformation to generate electricity load fluctuation data and curves that comprehensively depict the electricity consumption patterns of abnormal electricity meters. Dynamic monitoring capabilities help to promptly detect abnormal load fluctuations, laying the foundation for subsequent detection of illegal electricity use. By calculating the slope of the load fluctuation curve, the extracted load fluctuation feature data can quantify subtle changes in user electricity consumption behavior. Precise feature extraction improves the recognition rate of abnormal electricity consumption behavior, especially in complex or frequently fluctuating electricity consumption scenarios. Based on the load fluctuation feature data, illegal electricity use detection is performed on the abnormal electricity meter current data. The generated illegal electricity use detection data can effectively identify abnormal behaviors such as electricity theft and bypassing regulations. The intelligent processing of illegal electricity use detection significantly improves detection efficiency and accuracy, reducing the need for manual intervention. Peak-valley time period analysis of normal electricity consumption detection data generates normal electricity consumption behavior data that accurately reflects users' reasonable electricity consumption habits. Peak-valley time period analysis supports further optimization of power dispatch strategies and can also serve as a reference for electricity price design. Data on illegal electricity use generated through impact assessments can quantify the potential impact of illegal electricity use on the power grid and normal users. This quantitative analysis provides a scientific basis for subsequent sanctions and power grid optimization, and helps reduce the occurrence of illegal electricity use.

[0049] Preferably, step S4 includes the following steps:

[0050] Step S41: Divide the data on differences in the performance of electricity meter equipment and the data on differences in electricity consumption behavior of electricity meter into datasets to generate a model training set and a model test set; train the model on the model training set using the random forest algorithm to generate a pre-model for identifying differences in electricity meter current measurement.

[0051] Step S42: Use the model test set to perform model optimization and iteration on the pre-model for identifying differences in current metering of electricity meters, thereby generating a model for identifying differences in current metering of electricity meters; import the identification data for differences in the performance of electricity meter equipment and the identification data for differences in electricity consumption behavior of electricity meters into the model for identifying differences in current metering of electricity meters to classify differences in current metering of electricity meters, and generate classification matching data for differences in current metering of electricity meters.

[0052] Step S43: Visualize the data of the electricity meter current measurement difference classification and matching, and generate an electricity meter current measurement difference classification and matching report to perform automated operation of electricity meter current measurement difference identification.

[0053] This invention ensures a scientific separation of the training and testing sets through dataset partitioning, enabling the trained model to accurately capture the performance and electricity consumption behavior differences of electricity meters. The random forest algorithm is used for model training, offering powerful feature selection and classification accuracy, thus improving the accuracy of the electricity meter current measurement difference identification model. Iterative model optimization based on the test set continuously improves the robustness and generalization ability of the electricity meter current measurement difference identification model. The optimized model can more accurately classify and match current measurement differences, adapting to diverse electricity meter operating environments and abnormal behavior patterns. The integrated analysis of electricity meter performance difference data and electricity consumption behavior difference data provides multi-dimensional support for measurement difference identification and classification. Using the optimized model to complete the classification and matching of electricity meter current measurement differences can quickly and efficiently identify various types of differences, providing a reliable basis for subsequent decision-making. Data visualization presents the complex classification and matching results in an intuitive chart format, and the generated electricity meter current measurement difference classification and matching report clearly shows the difference distribution and key features. The visualization report supports automated operations, reducing the complexity of manual analysis and improving decision-making efficiency. By automating model training, classification, and report generation, the time cost and error risk of traditional manual processing are significantly reduced. This achieves full automation of the electricity meter metering discrepancy identification process, providing technical support for large-scale smart meter management and anomaly monitoring.

[0054] This specification provides a current metering difference identification system for electricity meters, used to execute the above-described current metering difference identification method for electricity meters. The current metering difference identification system includes:

[0055] The current difference discrimination module is used to acquire the location information data of the electricity meter; construct a current acquisition network for the electricity meter based on the location information data of the electricity meter, and generate a current acquisition network; acquire electricity meter current data from the current acquisition network to obtain standard electricity meter current acquisition data; and discriminate the current measurement difference of the standard electricity meter current acquisition data to generate electricity meter current measurement difference data.

[0056] The equipment analysis module is used to screen abnormal energy meters in the current acquisition network based on the current measurement difference data of the energy meters, thereby obtaining abnormal energy meter screening data; to perform current transformer performance testing on the abnormal energy meter screening data, generating abnormal energy meter hardware identification data; to verify the energy meter firmware version on the abnormal energy meter screening data, thereby generating abnormal energy meter software identification data; to perform electromagnetic interference analysis on the abnormal energy meter screening data, generating abnormal energy meter electromagnetic interference identification data; and to integrate the abnormal energy meter hardware identification data, abnormal energy meter software identification data, and abnormal energy meter electromagnetic interference identification data to generate energy meter equipment performance difference identification data.

[0057] The electricity consumption analysis module is used to analyze user electricity load fluctuations from abnormal electricity meter screening data and generate electricity load fluctuation curves; to detect illegal electricity use from the electricity load fluctuation curves and generate illegal electricity use detection data and normal electricity use detection data; and to perform stratified electricity use behavior impact analysis on the illegal electricity use detection data and normal electricity use detection data and generate electricity meter electricity use behavior difference identification data.

[0058] The difference identification module is used to train a model by using the difference identification data of electricity meter equipment performance and the difference identification data of electricity meter electricity consumption behavior to generate an electricity meter current measurement difference identification model. The difference identification data of electricity meter equipment performance and the difference identification data of electricity meter electricity consumption behavior are then imported into the electricity meter current measurement difference identification model to perform electricity meter current measurement difference identification and classification, and generate an electricity meter current measurement difference classification and matching report to perform automated electricity meter current measurement difference identification.

[0059] The beneficial effects of this invention lie in ensuring the accuracy of data acquisition by obtaining the location information of electricity meters and constructing a current acquisition network, thus providing high-quality standard current data for subsequent current metering difference discrimination. Current metering difference discrimination accurately identifies metering differences in electricity meters, quickly acquiring current metering difference data and providing a basis for equipment anomaly analysis and decision-making. The equipment analysis module, combined with current metering difference data, performs comprehensive screening and diagnosis of electricity meters, covering multiple levels of hardware, software, and environmental interference. Current transformer performance testing, firmware version verification, and electromagnetic interference analysis comprehensively improve the accuracy of equipment anomaly identification, helping to promptly detect hardware faults, software problems, and external interference. The integration of equipment performance difference identification data makes comprehensive diagnosis of electricity meter performance possible, providing more accurate information for fault diagnosis and optimization. The electricity consumption analysis module performs load fluctuation analysis on the abnormal electricity meter screening data, enabling the identification of irregular electricity consumption behavior. Through illegal electricity consumption detection and layered electricity consumption behavior impact analysis, it can accurately distinguish between normal and abnormal electricity consumption behavior, providing strong support for electricity meters to identify illegal electricity consumption behavior. The generation of electricity meter usage behavior difference identification data helps improve the accuracy of power supervision and effectively curb electricity theft. The difference identification module combines equipment performance differences with usage behavior differences, generating a current metering difference identification model through model training, ensuring high accuracy and efficiency in difference identification. The application of the electricity meter current metering difference identification model can accurately classify electricity meters and generate difference matching reports, further promoting the automation of electricity meter difference identification and reducing manual intervention. Through accurate current metering difference identification and equipment performance analysis, the system can promptly detect electricity meter faults, illegal electricity use, and environmental interference issues, ensuring the stable operation of the power system. Automated operations reduce the time and cost of manual analysis, improve overall power management efficiency, and help power companies and users achieve more efficient resource allocation and management. The modular design and comprehensive analysis capabilities of this system promote the development of electricity meter management towards intelligence and automation. Automated identification of current metering differences, comprehensive diagnosis of equipment performance, and intelligent analysis of usage behavior will greatly enhance the intelligence level of the power industry in equipment management, resource allocation, and safety monitoring. Therefore, this invention improves the accuracy and reliability of identifying differences in current meter readings by constructing a current acquisition network, performing multi-dimensional anomaly analysis, assessing the impact of user behavior, and training the model. Attached Figure Description

[0060] Figure 1 A schematic diagram of the steps in a method for identifying differences in current measurement in an electricity meter.

[0061] Figure 2 for Figure 1 A detailed flowchart illustrating the implementation steps of step S2.

[0062] Figure 3 for Figure 1 A detailed flowchart illustrating the implementation steps of step S3.

[0063] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0064] The technical method of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0065] Furthermore, the accompanying drawings are merely illustrative of the invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor methods and / or microcontroller methods.

[0066] It should be understood that although the terms "first," "second," etc., may be used herein to describe various units, these units should not be limited by these terms. These terms are used merely to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, a first unit may be referred to as a second unit, and similarly, a second unit may be referred to as a first unit. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0067] To achieve the above objectives, please refer to Figures 1 to 3 A method for identifying differences in current measurement in an electricity meter, the method comprising the following steps:

[0068] Step S1: Obtain the location information data of the electricity meter; construct a current acquisition network for the electricity meter based on the location information data of the electricity meter to generate a current acquisition network; collect the current data of the electricity meter from the current acquisition network to obtain the current acquisition data of the standard electricity meter; determine the difference in current measurement of the standard electricity meter from the current acquisition data to generate the difference in current measurement of the electricity meter.

[0069] Step S2: Based on the current measurement difference data of the electricity meters, the current acquisition network is screened for abnormal electricity meters to obtain abnormal electricity meter screening data; the performance of the current transformer is tested on the abnormal electricity meter screening data to generate abnormal electricity meter hardware identification data; the electricity meter firmware version is verified on the abnormal electricity meter screening data to generate abnormal electricity meter software identification data; electromagnetic interference analysis is performed on the abnormal electricity meter screening data to generate abnormal electricity meter electromagnetic interference identification data; the abnormal electricity meter hardware identification data, abnormal electricity meter software identification data, and abnormal electricity meter electromagnetic interference identification data are integrated to generate electricity meter equipment performance difference identification data.

[0070] Step S3: Perform user electricity load fluctuation analysis on the abnormal electricity meter screening data to generate electricity load fluctuation curves; perform illegal electricity use detection on the electricity load fluctuation curves to generate illegal electricity use detection data and normal electricity use detection data; perform hierarchical electricity use behavior impact analysis on the illegal electricity use detection data and normal electricity use detection data to generate electricity meter electricity use behavior difference identification data.

[0071] Step S4: Train the model using the data on differences in the performance of the electricity meter equipment and the data on differences in the electricity consumption behavior of the electricity meter to generate a model for identifying differences in the current measurement of the electricity meter; import the data on differences in the performance of the electricity meter equipment and the data on differences in the electricity consumption behavior of the electricity meter into the model for identifying and classifying differences in the current measurement of the electricity meter, and generate a classification and matching report for the difference in the current measurement of the electricity meter to perform automated operation of identifying differences in the current measurement of the electricity meter.

[0072] This invention constructs a location-based current acquisition network to ensure comprehensive current acquisition and network coverage accuracy. The standardized current data collected improves the accuracy of metering discrepancy identification and reduces the impact of data inconsistencies and environmental interference on discrepancy recognition. By screening abnormal energy meters, problematic devices can be quickly located. Combining hardware performance testing, firmware version verification, and electromagnetic interference analysis, the causes of anomalies are analyzed from multiple dimensions, providing a reliable basis for device performance optimization. The integrated device performance difference data helps accurately identify metering errors caused by hardware or software. Abnormal electricity usage patterns are identified through electricity load fluctuation curve analysis. Illegal electricity usage detection provides technical support for preventing electricity theft or irregular electricity use. Layered electricity usage behavior impact analysis refines the impact of user behavior on metering discrepancies, laying the foundation for comprehensive identification of problem causes. Combining device performance and user behavior difference data, model training improves the accuracy and reliability of metering discrepancy identification. Generating discrepancy identification reports through classification and matching automates and simplifies the identification process, reducing the cost and error of manual intervention. Therefore, this invention improves the accuracy and reliability of energy meter current metering discrepancy identification by constructing a current acquisition network, performing multi-dimensional anomaly analysis, assessing the impact of user behavior, and training a model.

[0073] In this embodiment of the invention, reference is made to Figure 1 The diagram shown is a flowchart illustrating the steps of a method for identifying current metering differences in an electricity meter according to the present invention. In this example, the method for identifying current metering differences in an electricity meter includes the following steps:

[0074] Step S1: Obtain the location information data of the electricity meter; construct a current acquisition network for the electricity meter based on the location information data of the electricity meter to generate a current acquisition network; collect the current data of the electricity meter from the current acquisition network to obtain the current acquisition data of the standard electricity meter; determine the difference in current measurement of the standard electricity meter from the current acquisition data to generate the difference in current measurement of the electricity meter.

[0075] In this embodiment of the invention, location information data of electricity meters is collected through a power management system database or IoT devices. Location information may include GPS coordinates, zone number, and transformer area number. If data is missing, the missing information is supplemented through geographic calculation or device association to generate complete electricity meter location information data. Based on the location information data, a logical network connection is established between electricity meter devices. A topology analysis method is used to design the hierarchical structure of the acquisition network according to the distribution of power transformer areas or the power grid, including: Main station: a central acquisition node used to aggregate regional current data. Substations: distributed acquisition nodes responsible for collecting current data within subnets or areas. Sub-nodes: specific electricity meter devices. The acquisition network is ensured to have redundancy and anti-interference capabilities, supporting real-time data acquisition and transmission. Current data from each electricity meter is collected in real time using current sensors or metering modules. The collected raw current data contains noise or outliers and requires the following processing: noise and outliers are removed. The current data is converted to a unified standard format (e.g., A, mA units). The acquisition cycle and time are synchronized to support the accuracy of subsequent difference discrimination, obtaining standard electricity meter current acquisition data. Set a current reference value (e.g., based on the meter specifications and actual load conditions). Calculate the difference between the collected data and the reference current, and determine the difference using the formula: Di = |Ii−Iref|, where Di is the current difference value of the i-th meter, Ii is the collected current, and Iref is the reference current. If the difference value Di exceeds a preset threshold, it is determined to be a metering discrepancy. The determination result is classified as "normal" or "abnormal," and a classification label is generated.

[0076] Step S2: Based on the current measurement difference data of the electricity meters, the current acquisition network is screened for abnormal electricity meters to obtain abnormal electricity meter screening data; the performance of the current transformer is tested on the abnormal electricity meter screening data to generate abnormal electricity meter hardware identification data; the electricity meter firmware version is verified on the abnormal electricity meter screening data to generate abnormal electricity meter software identification data; electromagnetic interference analysis is performed on the abnormal electricity meter screening data to generate abnormal electricity meter electromagnetic interference identification data; the abnormal electricity meter hardware identification data, abnormal electricity meter software identification data, and abnormal electricity meter electromagnetic interference identification data are integrated to generate electricity meter equipment performance difference identification data.

[0077] In this embodiment of the invention, abnormal energy meters are screened from the current acquisition network based on the abnormal classification tags marked in the current metering difference data. Screening criteria: current difference value Di exceeds a preset threshold; data fluctuations are large and cannot meet metering stability requirements. The screening results are output, marking the device number and location information of each abnormal energy meter, generating abnormal energy meter screening data, including a list of abnormal energy meters and their location information. Characteristic parameters of the current transformer of each abnormal energy meter are detected, including current ratio, phase offset, and saturation characteristics. The acquired current signal is subjected to spectral decomposition to detect harmonic components and determine whether the transformer performance meets the standard. Based on performance parameters and spectral characteristics, fault tree analysis is used to locate transformer hardware problems, generating abnormal energy meter hardware identification data, recording transformer performance and potential fault information. The firmware version information of the abnormal energy meters is obtained, and its compatibility and functionality with the latest version are analyzed. The presence of known vulnerabilities or functional defects in the firmware version is verified. Firmware that does not conform to the current standard or has known problems is marked, generating abnormal energy meter software identification data, including firmware version verification results and related defect information. Three-dimensional electromagnetic field scanning technology was used to obtain the electromagnetic field intensity distribution of the installation environment of abnormal energy meters. Wavelet transform was performed on the electromagnetic field intensity distribution data to extract the time-frequency characteristics of electromagnetic interference and generate interference spectrum data. Based on the interference spectrum data, the interference source was located using multi-point triangulation, and the interference propagation path and attenuation characteristics were analyzed. Electromagnetic field-circuit co-simulation technology was combined to simulate the coupling effect of electromagnetic interference on the energy meter, classify and identify interference modes, generate electromagnetic interference identification data for abnormal energy meters, and record the location, propagation characteristics, and coupling modes of the interference source. Correlation analysis was performed on the hardware, software, and electromagnetic interference identification data to establish a unified performance difference model. The integrated data was classified to clarify the performance problem category and impact degree of each abnormal energy meter, generating energy meter equipment performance difference identification data, providing a basis for subsequent optimization and anomaly handling.

[0078] Step S3: Perform user electricity load fluctuation analysis on the abnormal electricity meter screening data to generate electricity load fluctuation curves; perform illegal electricity use detection on the electricity load fluctuation curves to generate illegal electricity use detection data and normal electricity use detection data; perform hierarchical electricity use behavior impact analysis on the illegal electricity use detection data and normal electricity use detection data to generate electricity meter electricity use behavior difference identification data.

[0079] In this embodiment of the invention, the current consumption data of each abnormal energy meter within a specified time range is collected; the load fluctuation value on the time series is calculated: load fluctuation value at time t = |current value at time t - average current value of the time series| ÷ average current value of the time series. The load fluctuation value is plotted as a time series curve to generate an electricity load fluctuation curve, reflecting the user's electricity consumption fluctuation characteristics. The local slope of the electricity load fluctuation curve is calculated: load fluctuation slope at time t = (F(t+1)−F(t)) ÷ time interval. Abnormal slope values ​​and their distribution are identified to locate abnormal electricity consumption behavior. The load fluctuation characteristics are matched with a known illegal electricity consumption behavior pattern library to detect whether there is electricity theft, wire skipping, or other illegal behavior. The type, time point, and scope of impact of the detected illegal electricity consumption behavior are recorded. Normal electricity load information for which no abnormal behavior was found is recorded. Based on the illegal electricity consumption detection data, the impact of illegal behavior on electricity stability, metering accuracy, and equipment safety is analyzed to generate illegal electricity consumption behavior impact indicators, including electricity loss rate, energy meter abnormality rate, etc. The normal electricity consumption monitoring data is analyzed by time period stratification (peak and valley periods, day and night periods). The electricity load distribution and power factor characteristics of each time period are calculated to generate electricity meter electricity consumption behavior difference identification data, including: data on the impact of illegal electricity consumption behavior; characteristic data of normal electricity consumption periods; and classification results of electricity consumption behavior differences.

[0080] Step S4: Train the model using the data on differences in the performance of the electricity meter equipment and the data on differences in the electricity consumption behavior of the electricity meter to generate a model for identifying differences in the current measurement of the electricity meter; import the data on differences in the performance of the electricity meter equipment and the data on differences in the electricity consumption behavior of the electricity meter into the model for identifying and classifying differences in the current measurement of the electricity meter, and generate a classification and matching report for the difference in the current measurement of the electricity meter to perform automated operation of identifying differences in the current measurement of the electricity meter.

[0081] In this embodiment of the invention, the input data includes: electricity meter equipment performance difference identification data (including hardware performance, software performance, electromagnetic interference characteristics, etc.); and electricity meter electricity consumption behavior difference identification data (including load fluctuation characteristics, illegal electricity consumption behavior characteristics, normal electricity consumption behavior characteristics, etc.). The input data is randomly divided into a model training set and a model test set to ensure balanced data distribution and compliance with statistical laws. The random forest algorithm is used to train the training set data, extracting key features affecting electricity meter current measurement differences, and generating a preliminary pre-model for electricity meter current measurement difference identification. The test set is used to verify the model performance, evaluating the model's classification accuracy, recall, and specificity. Based on the test results, the model is iteratively optimized by adjusting the decision tree depth, feature weights, and data balancing parameters, ultimately generating an electricity meter current measurement difference identification model. The equipment performance difference identification data and electricity consumption behavior difference identification data are input into the optimized model; a classification operation is performed to identify and match current measurement difference categories, generating electricity meter current measurement difference classification matching data. The distribution of current metering differences across different categories is displayed using bar charts, heatmaps, and scatter plots. Multidimensional analysis of the difference data is performed based on region, equipment model, or time dimension to intuitively reflect the classification results and generate a classification and matching report of current metering differences for automated operation.

[0082] Preferably, step S1 includes the following steps:

[0083] Step S11: Obtain the location information data of the electricity meter;

[0084] Step S12: Construct a current acquisition network for the electricity meter based on the location information data of the electricity meter to generate a current acquisition network; collect electricity meter current data from the current acquisition network at the same time interval to obtain electricity meter current acquisition data.

[0085] Step S13: Perform data preprocessing on the current acquisition data of the electricity meter to generate standard current acquisition data of the electricity meter. The data preprocessing includes data cleaning, data denoising, data missing value filling and standardization.

[0086] Step S14: Analyze the current measurement differences of the standard electricity meter data, generate current measurement difference data and current measurement non-difference data, and remove the current measurement non-difference data.

[0087] In this embodiment of the invention, the latitude and longitude location information of each electricity meter is obtained through a Geographic Information System (GIS) module in the power management system; or the location information of the electricity meters is recorded through on-site inspection equipment (such as GPS terminals or mobile data acquisition devices). The electricity meter location information data is stored in a database, including key information such as the electricity meter ID, geographical location (latitude and longitude), area code, and installation environment. Based on the obtained electricity meter location information data, a current acquisition network is established using a topology algorithm. The nodes of the network are electricity meters, and the edges represent physical or logical connections between adjacent electricity meters. Weight values ​​are introduced to label the edges, calculated based on geographical distance, line type, or signal transmission delay. A uniform time interval (e.g., 1 minute or 5 minutes) is set to collect current data from each electricity meter in the current acquisition network in real time. The data includes timestamps, current values, phase information, etc., forming an initial dataset for electricity meter current acquisition. Abnormal data points (such as data with current values ​​of zero or exceeding a reasonable range) are detected and removed. Data integrity is compared, and missing timestamp information is supplemented. Low-pass filters or wavelet transforms are applied to remove random noise from the signal. Based on historical data trends, abnormal fluctuations are smoothed using mean or median filtering. Small-range missing values ​​are filled using interpolation algorithms (such as linear or Lagrange interpolation). Large-range missing data are filled using a machine learning-based regression prediction model. Current values ​​are normalized to the 0,1 range according to the maximum and minimum values ​​of each electricity meter. If the collection frequencies of different electricity meters are inconsistent, the data is time-aligned. The difference (absolute or relative difference) between the current collection value of each electricity meter and the current values ​​of its neighboring electricity meters in the collection network is calculated. Statistical discriminant algorithms (such as Z-Score or IQR) are applied to perform significance analysis on the differences, generating current measurement difference data. All data with insignificant measurement differences are marked as "non-discrepancy data." Non-discrepancy data is removed from the standard electricity meter current collection dataset to optimize the efficiency and accuracy of subsequent processing.

[0088] Preferably, the construction of a current acquisition network for electricity meters based on electricity meter location information data includes the following steps:

[0089] Geospatial mapping is performed on the location information data of electricity meters to generate spatial mapping data of electricity meter locations; based on the spatial mapping data of electricity meter locations, the current collection areas of electricity meters are divided to generate current collection area data of electricity meters.

[0090] Based on the data of the current collection area of ​​the electricity meter, the network topology is selected to obtain the network topology data of the electricity meter current; based on the network topology data of the electricity meter current, the regional electricity meter density is analyzed to generate regional electricity meter density data.

[0091] Based on the regional electricity meter density data, gateways are deployed for the electricity meter current acquisition area data to generate electricity meter acquisition gateway deployment data; the regional electricity meter density data is collected and transmitted centrally through the electricity meter acquisition gateway deployment data to generate a current acquisition network; electricity meter current data is collected from the current acquisition network at the same time intervals to obtain electricity meter current acquisition data.

[0092] In this embodiment of the invention, a geospatial distribution map is generated by loading the location information of electricity meters into a Geographic Information System (GIS). Spatial mapping algorithms (such as Voronoi diagrams or Delaunay triangulation) are used to divide the electricity meter locations into regions, forming spatial mapping data of electricity meter locations. Based on the geographical distribution of the electricity meters and their respective power supply areas, hierarchical clustering or density-based clustering (such as DBSCAN) is used to divide the current collection areas of the electricity meters. The division rules for each region can be based on the following criteria: geographical distance, regional load balance, and electricity meter quantity limits. For each electricity meter collection area, a suitable network topology (such as star, mesh, or hybrid topology) is designed. The network topology selection follows these principles: star topology is suitable for low-density areas and is easy to centrally manage; mesh topology is suitable for high-density areas and enhances the reliability and redundancy of data transmission. A weighted graph algorithm (such as minimum spanning tree or shortest path tree) is used to generate the network topology within the region. The regional electricity meter density (the number of electricity meters per unit area) is calculated for each current collection area. Statistical methods are used to generate density heatmaps, displaying the distribution of high and low density. Gateway deployment locations are determined based on the regional electricity meter density and network topology. Optimization algorithms (such as K-means clustering or integer programming) are used to cover all electricity meters with the fewest gateways possible, while simultaneously meeting data transmission latency and load balancing requirements. Current data from the deployed gateways is centrally collected, and data transmission paths are optimized to reduce latency. The collection period can be set to a uniform time interval (e.g., 1 minute or 5 minutes) to periodically transmit current data to the central server. Real-time current data is collected from each electricity meter within the same time interval through the current collection network. The collected data includes additional information such as current value, timestamp, voltage, and phase.

[0093] Preferably, step S14 includes the following steps:

[0094] The average current of the standard energy meter is calculated by analyzing the current data collected from the standard energy meter. The average current of the energy meter is then used as the reference current value, and the current measurement difference of the standard energy meter current data is calculated based on the reference current value to obtain the current measurement difference value.

[0095] The current metering difference value is judged according to the preset current metering difference threshold. When the current metering difference value is greater than or equal to the preset current metering difference threshold, the corresponding current metering difference value is marked as the current metering difference data of the energy meter.

[0096] When the current metering difference value is less than the preset current metering difference threshold, the corresponding current metering difference value is marked as non-difference data of the current meter, and the non-difference data of the current meter are removed from the data in the same time period.

[0097] In this embodiment of the invention, the average current is calculated by averaging the standard current data collected from each electricity meter (current data collected over a certain time interval). The calculation formula is: Average Current = in, For the first time interval, the data collected is... One current data point, This represents the number of data points collected within this time period. The average current of each electricity meter is used as the baseline current value for subsequent current metering difference calculations. For each electricity meter, the difference between the collected data and the corresponding average current is calculated to obtain the current metering difference value. The calculation formula is: Current metering difference value = |I_collected - I_average|, where I_collected is the current data collected each time, and I_average is the average current of the electricity meter. The current metering difference value of each electricity meter reflects the difference between the current collected data and the average current. Based on a preset current metering difference threshold, the magnitude of each current metering difference value is determined. If the current metering difference value is greater than or equal to the threshold, it is marked as electricity meter current metering difference data; if it is less than the threshold, it is marked as electricity meter current metering non-difference data. For data marked as electricity meter current metering non-difference, further data removal within the same time period is performed. Non-difference data within the same time period is marked as redundant data and removed in subsequent processing. The removal method can use timestamp-based filtering to delete redundant data within the same time period.

[0098] As an example of the present invention, reference is made to... Figure 2 As shown, in this example, step S2 includes:

[0099] Step S21: Screen abnormal energy meters in the current acquisition network based on the difference data of current meter readings, and obtain abnormal energy meter screening data.

[0100] Step S22: Perform current transformer performance testing on the abnormal energy meter screening data to generate energy meter current transformer performance data; based on the energy meter current transformer performance data, perform energy meter internal circuit testing on the abnormal energy meter screening data to generate abnormal energy meter hardware identification data.

[0101] Step S23: Check the calibration parameters of the abnormal energy meter screening data to obtain the abnormal energy meter parameter calibration record; verify the energy meter firmware version of the abnormal energy meter screening data through the abnormal energy meter parameter calibration record to generate abnormal energy meter software identification data.

[0102] Step S24: Monitor the working environment of the abnormal energy meter screening data to generate abnormal energy meter environmental monitoring data; perform electromagnetic interference analysis on the abnormal energy meter environmental monitoring data to generate abnormal energy meter electromagnetic interference identification data; integrate the abnormal energy meter hardware identification data, abnormal energy meter software identification data, and abnormal energy meter electromagnetic interference identification data to generate energy meter equipment performance difference identification data.

[0103] In this embodiment of the invention, by screening based on current measurement difference data, energy meters with current measurement difference values ​​greater than or equal to a preset threshold are first identified. If the current measurement difference value exceeds the preset threshold, the energy meter is considered abnormal and requires further processing. These energy meters are marked as "abnormal energy meters," generating abnormal energy meter screening data. Current transformer performance testing is performed on the screened abnormal energy meters to check the accuracy and working status of the transformers. Specific methods include comparing with standard transformers and measuring indicators such as current ratio, phase difference, and load capacity. The test results generate energy meter current transformer performance data to assess whether there are performance problems with the transformers. Circuit testing is performed on the screened abnormal energy meters to check the function and performance of hardware components such as circuit boards, connectors, and sensors. The system checks for short circuits, open circuits, or other hardware faults to ensure accurate current data acquisition. The test results generate abnormal energy meter hardware identification data. Calibration parameters are checked on the abnormal energy meters to verify whether the meter calibration meets the prescribed standards. The difference between the actual measurement results and the expected standard values ​​is compared to generate abnormal energy meter parameter calibration records. Firmware version verification is performed on the abnormal energy meters to check if the firmware version is up-to-date and matches the calibration requirements. If a known issue exists in the firmware version affecting current measurement results, abnormal energy meter software identification data is generated to indicate whether it is a software fault or firmware problem. The operating environment of the abnormal energy meters is monitored, including factors such as temperature, humidity, and power quality, to confirm whether the external environment meets the operating requirements of the energy meters. Abnormal energy meter environmental monitoring data is generated to help determine whether environmental factors affect the energy meter performance. Electromagnetic interference analysis is performed on the abnormal energy meters to check whether external electromagnetic interference affects the accuracy of current measurement. If electromagnetic interference causes misreading or deviation of the energy meters, it needs to be analyzed to ensure the reliability of the current data. Abnormal energy meter electromagnetic interference identification data is generated to identify the energy meters affected by electromagnetic interference. Input data: Abnormal energy meter hardware identification data (from step S22), abnormal energy meter software identification data (from step S23), abnormal energy meter electromagnetic interference identification data (from step S24). The above three types of data are integrated to form comprehensive energy meter device performance difference identification data. Based on the results of hardware, software, and environmental analysis, determine the specific reasons for the differences in equipment performance, such as whether they are caused by factors such as current transformer failure, firmware issues, or electromagnetic interference.

[0104] Preferably, step S22 includes the following steps:

[0105] Step S221: Perform linear feature analysis on the abnormal energy meter screening data to generate current transformer characteristic parameter data; perform spectral decomposition on the current transformer characteristic parameter data to generate current harmonic component data.

[0106] Step S222: Perform dynamic response modeling on the current harmonic component data to generate the current transformer performance data of the energy meter; perform fault tree analysis on the current transformer performance data of the energy meter to generate circuit node status data.

[0107] Step S223: Perform circuit simulation on the circuit node state data to generate circuit operation characteristic data; perform pattern recognition on the abnormal energy meter screening data based on the circuit operation characteristic data to generate abnormal energy meter hardware identification data.

[0108] In this embodiment of the invention, linear feature analysis is performed on the current data of abnormal energy meters to extract linear features from the current signal, such as amplitude, phase, and frequency. Characteristic parameter data of the current transformer is obtained through mathematical models or statistical analysis methods, such as least squares method and principal component analysis (PCA). These parameters help identify the operating state of the current transformer, generating current transformer characteristic parameter data, which includes linear feature data of the current signal. The extracted current transformer characteristic parameter data is then subjected to spectral decomposition using methods such as Fourier transform or wavelet transform to decompose the current signal into harmonic components of different frequencies. Each harmonic component is analyzed to identify abnormal frequency characteristics, indicating a problem with the current transformer, such as harmonic distortion. Current harmonic component data is generated, including the frequency, amplitude, and relationship of each harmonic component to the fundamental current. Dynamic response modeling is then performed on the current harmonic component data to simulate the dynamic response of the current transformer under different operating states, including harmonic transmission and filtering characteristics. Control theory or signal processing models can be used to reconstruct the performance of current transformers during operation through time and frequency domain analysis, generating performance data for the current transformers in the energy meter. This data includes the frequency response, dynamic behavior, and responsiveness to different current loads. Based on the generated performance data, fault tree analysis is performed to determine the potential causes of current transformer failure. By analyzing failure modes (such as electromagnetic interference and current overload), a corresponding fault tree is generated, leading to the derivation of fault states at different nodes in the circuit, generating circuit node state data, and identifying faulty nodes or components. Based on the circuit node state data, the energy meter circuit is simulated to demonstrate its operation under different fault states. Circuit simulation software (such as SPICE) can be used to simulate factors such as current paths, power losses, and overloads, determining the impact of circuit faults and generating circuit operating characteristic data. This data describes the circuit's response characteristics under various operating environments, including voltage and current waveforms. Based on the circuit operating characteristic data, hardware identification of abnormal energy meters is performed using pattern recognition algorithms (such as support vector machines and neural networks). By comparing the circuit's operating characteristics with known normal operating modes, fault modes that differ from the standard mode are identified, abnormal energy meter hardware identification data is generated, and the specific type of energy meter hardware fault is identified, such as current transformer fault, circuit board problem, etc.

[0109] Preferably, electromagnetic interference analysis of abnormal electricity meter environmental monitoring data includes:

[0110] Three-dimensional electromagnetic field scanning and mapping are performed on the environmental monitoring data of abnormal energy meters to generate electromagnetic field intensity distribution data; wavelet transform time-frequency decomposition is performed on the electromagnetic field intensity distribution data to generate electromagnetic interference spectrum data.

[0111] Multi-point triangulation is performed on the electromagnetic interference spectrum data to locate the electromagnetic interference source; ray tracing attenuation analysis is performed on the electromagnetic interference source location data to generate electromagnetic interference propagation characteristic data.

[0112] Electromagnetic field-circuit co-simulation was performed on electromagnetic interference propagation characteristic data to generate electromagnetic interference coupling mode data.

[0113] The electromagnetic interference coupling mode data is processed by neural network classification and recognition to generate electromagnetic interference identification data for abnormal energy meters.

[0114] In this embodiment of the invention, a three-dimensional electromagnetic field scanning instrument or sensor is used to scan the electromagnetic field distribution of the electricity meter and its surrounding environment. The electromagnetic field in the environment is mapped in three-dimensional space to obtain the intensity data of the electromagnetic field at various spatial locations, generating electromagnetic field intensity distribution data. This data represents the electromagnetic field intensity at different locations, providing preliminary information about the source of electromagnetic interference. Wavelet transform is performed on the electromagnetic field intensity distribution data to extract electromagnetic field signal information in different frequency bands. Wavelet transform can provide high-precision local information between the time and frequency domains, enabling the separation of electromagnetic interference components within different frequency ranges, generating electromagnetic interference spectrum data. This data, through time-frequency decomposition, shows the variation characteristics of the electromagnetic interference signal under different frequencies and time windows. A multi-point triangulation method is used to compare the spectrum data of multiple electromagnetic field detection points with known locations to calculate the accurate location of the electromagnetic interference source. Based on the measured electromagnetic field intensity, combined with a geometric model, the interference source is located using the triangulation principle, generating electromagnetic interference source location data and identifying the spatial coordinates of the electromagnetic interference source. A ray tracing algorithm is used to simulate the propagation path of electromagnetic waves in the environment based on the spatial location data of the electromagnetic interference source. This study calculates the attenuation effect during electromagnetic wave propagation, considering the influence of environmental obstacles and electrical equipment. Electromagnetic interference (EMI) propagation characteristic data is generated to describe the propagation path, attenuation characteristics, and intensity distribution of EMI waves from the source to the target area. The EMI propagation characteristic data is then co-simulated with a circuit model of an energy meter to analyze the impact of EMI on the meter's internal circuitry. The simulation considers the coupling effect of electromagnetic fields on the circuit, simulating the interaction between EMI waves and various components in the circuit, particularly the responses of current transformers and processing circuits. This generates EMI coupling mode data, characterizing the coupling mode between EMI and the internal circuitry of the energy meter and its impact on the meter's performance. Based on known EMI modes, a neural network is used for classification and identification. Inputting EMI coupling mode data, the neural network is trained and learns the characteristics of EMI's impact on the energy meter, identifying different types of EMI sources and interference modes, generating abnormal energy meter EMI identification data, marking the EMI type, and indicating whether the energy meter is affected by EMI.

[0115] As an example of the present invention, reference is made to... Figure 3 As shown, step S3 in this example includes:

[0116] Step S31: Collect user electricity consumption data from the abnormal energy meter screening data to obtain abnormal energy meter current data; perform load fluctuation analysis on the abnormal energy meter current data to generate electricity load fluctuation data; perform curve conversion on the electricity load fluctuation data to generate electricity load fluctuation curve.

[0117] Step S32: Calculate the slope of the power load fluctuation curve to obtain load fluctuation characteristic data; use the load fluctuation characteristic data to detect illegal electricity use of abnormal power meter current data, and generate illegal electricity use detection data and normal electricity use detection data.

[0118] Step S33: Based on the illegal electricity use detection data, conduct an impact assessment on the abnormal electricity meter current data to generate illegal electricity use behavior data; conduct peak and valley period electricity use behavior analysis on the normal electricity use detection data to generate normal electricity use behavior data.

[0119] Step S34: Integrate the illegal electricity use behavior data and the normal electricity use behavior data to generate electricity meter electricity use behavior difference identification data.

[0120] In this embodiment of the invention, the data acquisition module of a smart energy meter records the user's electricity current data at different time periods, generating abnormal energy meter electricity current data for subsequent load fluctuation analysis. Time series analysis technology is used to detect the fluctuation characteristics of the electricity current data. The fluctuation patterns of daily electricity load are analyzed, and peak, trough, and abrupt changes in electricity consumption are identified, generating electricity load fluctuation data and marking the load fluctuation characteristics at different time periods. The load fluctuation data is mathematically modeled, and polynomial fitting or interpolation algorithms are used to transform discrete data into a continuous electricity load fluctuation curve, generating an electricity load fluctuation curve to support subsequent feature analysis. The slope of the fluctuation curve is calculated to extract the load change rate at different time points. By analyzing the absolute value and trend of the slope, the characteristics of electricity consumption behavior are identified, generating load fluctuation feature data to characterize the dynamic characteristics of user electricity consumption. The load fluctuation feature data is compared with a normal electricity consumption model to identify abnormal behaviors, such as sudden high load increases or high load behavior during off-peak hours. Machine learning classification algorithms (such as support vector machines or neural networks) are used to further classify electricity usage behavior, distinguishing between illegal and normal electricity use, generating illegal and normal electricity use detection data, and recording suspected illegal electricity use behavior and normal electricity use patterns respectively. The impact of illegal electricity use behavior on the overall power grid is assessed, including power supply imbalance and energy loss caused by illegal electricity use. The severity of illegal electricity use behavior is quantified, calculating the electricity consumption and time periods involved, generating illegal electricity use behavior data, and recording the specific characteristics and impacts of illegal electricity use. Based on the peak and off-peak periods of electricity use data, user electricity usage habits during peak and off-peak hours are analyzed. Combined with load fluctuation characteristic data, it is assessed whether users exhibit energy-saving behaviors or abnormal electricity use patterns, generating normal electricity use behavior data and providing characteristic analysis results of user electricity use habits. Illegal electricity use behavior data is compared with normal electricity use behavior data, and correlation analysis is used to identify the characteristic differences of abnormal electricity use behavior. Using data fusion algorithms (such as Bayesian inference or weighted average), the two types of data are combined to generate difference identification results, generating electricity meter usage behavior difference identification data, which comprehensively describes the characteristic differences between illegal and normal electricity usage behavior.

[0121] Preferably, step S4 includes the following steps:

[0122] Step S41: Divide the data on differences in the performance of electricity meter equipment and the data on differences in electricity consumption behavior of electricity meter into datasets to generate a model training set and a model test set; train the model on the model training set using the random forest algorithm to generate a pre-model for identifying differences in electricity meter current measurement.

[0123] Step S42: Use the model test set to perform model optimization and iteration on the pre-model for identifying differences in current metering of electricity meters, thereby generating a model for identifying differences in current metering of electricity meters; import the identification data for differences in the performance of electricity meter equipment and the identification data for differences in electricity consumption behavior of electricity meters into the model for identifying differences in current metering of electricity meters to classify differences in current metering of electricity meters, and generate classification matching data for differences in current metering of electricity meters.

[0124] Step S43: Visualize the data of the electricity meter current measurement difference classification and matching, and generate an electricity meter current measurement difference classification and matching report to perform automated operation of electricity meter current measurement difference identification.

[0125] In this embodiment of the invention, input data is randomly sampled and divided into a model training set and a model test set according to a certain ratio (e.g., 8:2). This ensures the representativeness of the training and test sets, including equipment performance characteristic data, electricity consumption behavior data, and relevant classification labels. A Random Forest algorithm is used to learn and model the training data. Leveraging the characteristics of Random Forest, multiple decision trees are constructed and ensemble learning is performed to generate the ability to classify and predict current metering differences. Random Forest parameters (e.g., number of trees, maximum depth, split criteria) are optimized using grid search or cross-validation to improve model performance, generating a pre-model for identifying current metering differences in electricity meters, possessing preliminary classification capabilities. The pre-model is evaluated using the test set, calculating model performance metrics (e.g., accuracy, recall, F1 score). Based on the difference between the prediction results on the test set and the actual labels, model parameters are adjusted using optimization algorithms (e.g., gradient boosting or iterative reweighting). This optimization iteration is repeated until the model performance metrics reach a set threshold, generating the final model for identifying current metering differences in electricity meters, possessing high-precision classification capabilities. The input data is imported into the recognition model to classify the current measurement discrepancies of electricity meters. The classification process matches the type of current measurement discrepancy (such as hardware failure, illegal electricity use, environmental interference, etc.) based on the model's output, generating classification matching data for electricity meter current measurement discrepancies, providing a foundation for subsequent visualization and automation. Data visualization tools (such as Matplotlib, Power BI, Tableau) are used to generate intuitive classification matching charts, including: bar charts or pie charts of classification distribution, time-series analysis curves of various measurement discrepancies, heat maps of regional or equipment distribution, and visualization results that help quickly identify areas or equipment types with abnormal concentrations, generating electricity meter current measurement discrepancy classification matching charts. A standardized report is automatically generated, containing the following: a summary table of electricity meter current measurement discrepancy classification results, visualization analysis charts, key conclusions and recommendations from the classification results (such as recommendations to replace faulty equipment or check areas of illegal electricity use), generating an electricity meter current measurement discrepancy classification matching report to support automated decision-making in electricity meter maintenance and management.

[0126] This specification provides a current metering difference identification system for electricity meters, used to execute the above-described current metering difference identification method for electricity meters. The current metering difference identification system includes:

[0127] The current difference discrimination module is used to acquire the location information data of the electricity meter; construct a current acquisition network for the electricity meter based on the location information data of the electricity meter, and generate a current acquisition network; acquire electricity meter current data from the current acquisition network to obtain standard electricity meter current acquisition data; and discriminate the current measurement difference of the standard electricity meter current acquisition data to generate electricity meter current measurement difference data.

[0128] The equipment analysis module is used to screen abnormal energy meters in the current acquisition network based on the current measurement difference data of the energy meters, thereby obtaining abnormal energy meter screening data; to perform current transformer performance testing on the abnormal energy meter screening data, generating abnormal energy meter hardware identification data; to verify the energy meter firmware version on the abnormal energy meter screening data, thereby generating abnormal energy meter software identification data; to perform electromagnetic interference analysis on the abnormal energy meter screening data, generating abnormal energy meter electromagnetic interference identification data; and to integrate the abnormal energy meter hardware identification data, abnormal energy meter software identification data, and abnormal energy meter electromagnetic interference identification data to generate energy meter equipment performance difference identification data.

[0129] The electricity consumption analysis module is used to analyze user electricity load fluctuations from abnormal electricity meter screening data and generate electricity load fluctuation curves; to detect illegal electricity use from the electricity load fluctuation curves and generate illegal electricity use detection data and normal electricity use detection data; and to perform stratified electricity use behavior impact analysis on the illegal electricity use detection data and normal electricity use detection data and generate electricity meter electricity use behavior difference identification data.

[0130] The difference identification module is used to train a model by using the difference identification data of electricity meter equipment performance and the difference identification data of electricity meter electricity consumption behavior to generate an electricity meter current measurement difference identification model. The difference identification data of electricity meter equipment performance and the difference identification data of electricity meter electricity consumption behavior are then imported into the electricity meter current measurement difference identification model to perform electricity meter current measurement difference identification and classification, and generate an electricity meter current measurement difference classification and matching report to perform automated electricity meter current measurement difference identification.

[0131] The beneficial effects of this invention lie in ensuring the accuracy of data acquisition by obtaining the location information of electricity meters and constructing a current acquisition network, thus providing high-quality standard current data for subsequent current metering difference discrimination. Current metering difference discrimination accurately identifies metering differences in electricity meters, quickly acquiring current metering difference data and providing a basis for equipment anomaly analysis and decision-making. The equipment analysis module, combined with current metering difference data, performs comprehensive screening and diagnosis of electricity meters, covering multiple levels of hardware, software, and environmental interference. Current transformer performance testing, firmware version verification, and electromagnetic interference analysis comprehensively improve the accuracy of equipment anomaly identification, helping to promptly detect hardware faults, software problems, and external interference. The integration of equipment performance difference identification data makes comprehensive diagnosis of electricity meter performance possible, providing more accurate information for fault diagnosis and optimization. The electricity consumption analysis module performs load fluctuation analysis on the abnormal electricity meter screening data, enabling the identification of irregular electricity consumption behavior. Through illegal electricity consumption detection and layered electricity consumption behavior impact analysis, it can accurately distinguish between normal and abnormal electricity consumption behavior, providing strong support for electricity meters to identify illegal electricity consumption behavior. The generation of electricity meter usage behavior difference identification data helps improve the accuracy of power supervision and effectively curb electricity theft. The difference identification module combines equipment performance differences with usage behavior differences, generating a current metering difference identification model through model training, ensuring high accuracy and efficiency in difference identification. The application of the electricity meter current metering difference identification model can accurately classify electricity meters and generate difference matching reports, further promoting the automation of electricity meter difference identification and reducing manual intervention. Through accurate current metering difference identification and equipment performance analysis, the system can promptly detect electricity meter faults, illegal electricity use, and environmental interference issues, ensuring the stable operation of the power system. Automated operations reduce the time and cost of manual analysis, improve overall power management efficiency, and help power companies and users achieve more efficient resource allocation and management. The modular design and comprehensive analysis capabilities of this system promote the development of electricity meter management towards intelligence and automation. Automated identification of current metering differences, comprehensive diagnosis of equipment performance, and intelligent analysis of usage behavior will greatly enhance the intelligence level of the power industry in equipment management, resource allocation, and safety monitoring. Therefore, this invention improves the accuracy and reliability of identifying differences in current meter readings by constructing a current acquisition network, performing multi-dimensional anomaly analysis, assessing the impact of user behavior, and training the model.

[0132] Therefore, the embodiments should be considered as exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of the equivalents of the application are intended to be included within the invention.

[0133] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features of the invention herein.

Claims

1. A method for identifying differences in current measurement in an electricity meter, characterized in that, Includes the following steps: Step S1: Obtain the location information data of the electricity meter; A current acquisition network is constructed based on the location information data of the electricity meter, and a current acquisition network is generated. The current acquisition network is used to acquire current data from electricity meters to obtain standard electricity meter current acquisition data. The current measurement difference of standard electricity meters is judged based on the current acquisition data of standard electricity meters, and the current measurement difference data of electricity meters is generated. Step S2: Screen abnormal energy meters in the current acquisition network based on the difference data of current meter readings, and obtain abnormal energy meter screening data; The current transformer performance is tested on the screening data of abnormal energy meters to generate hardware identification data of abnormal energy meters. The abnormal energy meter screening data is verified by the energy meter firmware version to generate abnormal energy meter software identification data; electromagnetic interference analysis is performed on the abnormal energy meter screening data to generate abnormal energy meter electromagnetic interference identification data; the abnormal energy meter hardware identification data, abnormal energy meter software identification data and abnormal energy meter electromagnetic interference identification data are integrated to generate energy meter device performance difference identification data. Step S3: Analyze the user's electricity load fluctuations from the abnormal electricity meter screening data to generate an electricity load fluctuation curve; detect illegal electricity use from the electricity load fluctuation curve to generate illegal electricity use detection data and normal electricity use detection data. A stratified analysis of the impact of illegal electricity use detection data and normal electricity use detection data was conducted to generate data for identifying differences in electricity meter usage behavior. Step S4: Train the model using the data on differences in the performance of the electricity meter equipment and the data on differences in the electricity consumption behavior of the electricity meter to generate a model for identifying differences in the current measurement of the electricity meter. The data on differences in the performance of electricity meters and the data on differences in electricity consumption behavior are imported into the electricity meter current measurement difference identification model to classify and identify differences in electricity meter current measurement, and an electricity meter current measurement difference classification and matching report is generated to perform automated electricity meter current measurement difference identification.

2. The method for identifying differences in current measurement in an electricity meter according to claim 1, characterized in that, Step S1 includes the following steps: Step S11: Obtain the location information data of the electricity meter; Step S12: Construct a current acquisition network for the electricity meter based on the location information data of the electricity meter to generate a current acquisition network; collect electricity meter current data from the current acquisition network at the same time interval to obtain electricity meter current acquisition data. Step S13: Perform data preprocessing on the current acquisition data of the electricity meter to generate standard current acquisition data of the electricity meter. The data preprocessing includes data cleaning, data denoising, data missing value filling and standardization. Step S14: Analyze the current measurement differences of the standard electricity meter data, generate current measurement difference data and current measurement non-difference data, and remove the current measurement non-difference data.

3. The method for identifying differences in current measurement in an electricity meter according to claim 2, characterized in that, The construction of a current acquisition network for electricity meters based on their location information includes the following steps: Geospatial mapping is performed on the location information data of electricity meters to generate spatial mapping data of electricity meter locations; based on the spatial mapping data of electricity meter locations, the current collection areas of electricity meters are divided to generate current collection area data of electricity meters. Based on the data of the current collection area of ​​the electricity meter, the network topology is selected to obtain the network topology data of the electricity meter current; based on the network topology data of the electricity meter current, the regional electricity meter density is analyzed to generate regional electricity meter density data. Based on the regional electricity meter density data, gateways are deployed for the electricity meter current acquisition area data to generate electricity meter acquisition gateway deployment data; the regional electricity meter density data is collected and transmitted centrally through the electricity meter acquisition gateway deployment data to generate a current acquisition network; electricity meter current data is collected from the current acquisition network at the same time intervals to obtain electricity meter current acquisition data.

4. The method for identifying differences in current measurement in an electricity meter according to claim 2, characterized in that, Step S14 includes the following steps: The average current of the standard energy meter is calculated by analyzing the current data collected from the standard energy meter. The average current of the energy meter is then used as the reference current value, and the current measurement difference of the standard energy meter current data is calculated based on the reference current value to obtain the current measurement difference value. The current metering difference value is judged according to the preset current metering difference threshold. When the current metering difference value is greater than or equal to the preset current metering difference threshold, the corresponding current metering difference value is marked as the current metering difference data of the energy meter. When the current metering difference value is less than the preset current metering difference threshold, the corresponding current metering difference value is marked as non-difference data of the current meter, and the non-difference data of the current meter are removed from the data in the same time period.

5. The method for identifying differences in current measurement in an electricity meter according to claim 1, characterized in that, Step S2 includes the following steps: Step S21: Screen abnormal energy meters in the current acquisition network based on the difference data of current meter readings, and obtain abnormal energy meter screening data. Step S22: Perform current transformer performance testing on the abnormal energy meter screening data to generate energy meter current transformer performance data; based on the energy meter current transformer performance data, perform energy meter internal circuit testing on the abnormal energy meter screening data to generate abnormal energy meter hardware identification data. Step S23: Check the calibration parameters of the abnormal energy meter screening data to obtain the abnormal energy meter parameter calibration record; verify the energy meter firmware version of the abnormal energy meter screening data through the abnormal energy meter parameter calibration record to generate abnormal energy meter software identification data. Step S24: Monitor the working environment of the abnormal energy meter screening data to generate abnormal energy meter environmental monitoring data; perform electromagnetic interference analysis on the abnormal energy meter environmental monitoring data to generate abnormal energy meter electromagnetic interference identification data; integrate the abnormal energy meter hardware identification data, abnormal energy meter software identification data, and abnormal energy meter electromagnetic interference identification data to generate energy meter equipment performance difference identification data.

6. The method for identifying differences in current measurement in an electricity meter according to claim 4, characterized in that, Step S22 includes the following steps: Step S221: Perform linear feature analysis on the abnormal energy meter screening data to generate current transformer characteristic parameter data; perform spectral decomposition on the current transformer characteristic parameter data to generate current harmonic component data. Step S222: Perform dynamic response modeling on the current harmonic component data to generate the current transformer performance data of the energy meter; perform fault tree analysis on the current transformer performance data of the energy meter to generate circuit node status data. Step S223: Perform circuit simulation on the circuit node state data to generate circuit operation characteristic data; perform pattern recognition on the abnormal energy meter screening data based on the circuit operation characteristic data to generate abnormal energy meter hardware identification data.

7. The method for identifying differences in current measurement in an electricity meter according to claim 4, characterized in that, Electromagnetic interference analysis of abnormal electricity meter environmental monitoring data includes: Three-dimensional electromagnetic field scanning and mapping are performed on the environmental monitoring data of abnormal energy meters to generate electromagnetic field intensity distribution data; wavelet transform time-frequency decomposition is performed on the electromagnetic field intensity distribution data to generate electromagnetic interference spectrum data. Multi-point triangulation is performed on the electromagnetic interference spectrum data to locate the electromagnetic interference source; ray tracing attenuation analysis is performed on the electromagnetic interference source location data to generate electromagnetic interference propagation characteristic data. Electromagnetic field-circuit co-simulation was performed on electromagnetic interference propagation characteristic data to generate electromagnetic interference coupling mode data. The electromagnetic interference coupling mode data is processed by neural network classification and recognition to generate electromagnetic interference identification data for abnormal energy meters.

8. The method for identifying differences in current measurement in an electricity meter according to claim 1, characterized in that, Step S3 includes the following steps: Step S31: Collect user electricity consumption data from the abnormal energy meter screening data to obtain abnormal energy meter current data; perform load fluctuation analysis on the abnormal energy meter current data to generate electricity load fluctuation data; perform curve conversion on the electricity load fluctuation data to generate electricity load fluctuation curve. Step S32: Calculate the slope of the power load fluctuation curve to obtain load fluctuation characteristic data; use the load fluctuation characteristic data to detect illegal electricity use of abnormal power meter current data, and generate illegal electricity use detection data and normal electricity use detection data. Step S33: Based on the illegal electricity use detection data, conduct an impact assessment on the abnormal electricity meter current data to generate illegal electricity use behavior data; conduct peak and valley period electricity use behavior analysis on the normal electricity use detection data to generate normal electricity use behavior data. Step S34: Integrate the illegal electricity use behavior data and the normal electricity use behavior data to generate electricity meter electricity use behavior difference identification data.

9. The method for identifying differences in current measurement in an electricity meter according to claim 1, characterized in that, Step S4 includes the following steps: Step S41: Divide the data on differences in the performance of electricity meter equipment and the data on differences in electricity consumption behavior of electricity meter into datasets to generate a model training set and a model test set; train the model on the model training set using the random forest algorithm to generate a pre-model for identifying differences in electricity meter current measurement. Step S42: Use the model test set to perform model optimization and iteration on the pre-model for identifying differences in current metering of electricity meters, thereby generating a model for identifying differences in current metering of electricity meters; import the identification data for differences in the performance of electricity meter equipment and the identification data for differences in electricity consumption behavior of electricity meters into the model for identifying differences in current metering of electricity meters to classify differences in current metering of electricity meters, and generate classification matching data for differences in current metering of electricity meters. Step S43: Visualize the data of the electricity meter current measurement difference classification and matching, and generate an electricity meter current measurement difference classification and matching report to perform automated operation of electricity meter current measurement difference identification.

10. A current metering difference identification system for electricity meters, characterized in that, For performing the electricity meter current measurement difference identification method as described in claim 1, the electricity meter current measurement difference identification system includes: The current difference discrimination module is used to acquire the location information data of the electricity meter; construct a current acquisition network for the electricity meter based on the location information data of the electricity meter, and generate a current acquisition network; acquire electricity meter current data from the current acquisition network to obtain standard electricity meter current acquisition data; and discriminate the current measurement difference of the standard electricity meter current acquisition data to generate electricity meter current measurement difference data. The equipment analysis module is used to screen abnormal energy meters in the current acquisition network based on the current measurement difference data of the energy meters, thereby obtaining abnormal energy meter screening data; to perform current transformer performance testing on the abnormal energy meter screening data, generating abnormal energy meter hardware identification data; to verify the energy meter firmware version on the abnormal energy meter screening data, thereby generating abnormal energy meter software identification data; to perform electromagnetic interference analysis on the abnormal energy meter screening data, generating abnormal energy meter electromagnetic interference identification data; and to integrate the abnormal energy meter hardware identification data, abnormal energy meter software identification data, and abnormal energy meter electromagnetic interference identification data to generate energy meter equipment performance difference identification data. The electricity consumption analysis module is used to analyze user electricity load fluctuations from abnormal electricity meter screening data and generate electricity load fluctuation curves; to detect illegal electricity use from the electricity load fluctuation curves and generate illegal electricity use detection data and normal electricity use detection data; and to perform stratified electricity use behavior impact analysis on the illegal electricity use detection data and normal electricity use detection data and generate electricity meter electricity use behavior difference identification data. The difference identification module is used to train a model by using the difference identification data of electricity meter equipment performance and the difference identification data of electricity meter electricity consumption behavior to generate an electricity meter current measurement difference identification model. The difference identification data of electricity meter equipment performance and the difference identification data of electricity meter electricity consumption behavior are then imported into the electricity meter current measurement difference identification model to perform electricity meter current measurement difference identification and classification, and generate an electricity meter current measurement difference classification and matching report to perform automated electricity meter current measurement difference identification.

Images