An Abnormal Location Method and System Based on Meter Reading Data

Abstract

The present invention discloses an abnormal positioning method and system based on meter reading data, which relates to the technical field of power system data processing and analysis, and includes: collecting multi-source data, performing fusion processing on the collected multi-source data, and subjecting the fused data to standardization processing to obtain a power system state matrix, identifying abnormal situations in the power system, quantifying the degree of abnormality of each node, judging whether the meter reading data is abnormal based on an anomaly detection value model, constructing an intelligent decision-making model according to historical data and current anomaly detection results, evaluating the comprehensive decision-making score of the node, judging the priority of the node based on the scoring result, matching maintenance tasks, judging the inaccessibility of the node, generating a processing instruction, and encrypting and storing the meter reading data. It enhances the ability of anomaly detection and fault warning, and improves the reliability, safety and maintenance efficiency of the power system operation.

Description

Technical Field

[0001] The present invention relates to the technical field of power system data processing and analysis, and specifically to an abnormal location method and system based on meter reading data. Background Art

[0002] With the rapid development of smart grid technology, the power meter reading system has gradually transitioned from traditional manual meter reading to the stage of automated and intelligent meter reading. The traditional manual meter reading method has low efficiency and poor accuracy, and is prone to human errors. These devices can automatically collect users' electricity consumption data and transmit the data to the power station through a wireless network. However, with the increase in data volume and the complexity of the network environment, problems such as data loss, delay, and data security during data transmission have become increasingly prominent. At the same time, due to the real-time and high-frequency nature of meter reading data, how to efficiently process and store this data, especially how to ensure data security and integrity, has become an urgent problem in the current power system.

[0003] Existing meter reading data processing systems have some deficiencies in data collection, storage, and analysis. First, although smart meters and sensors can provide high-precision electricity consumption data, there are still problems of data loss and difficulty in timely identifying abnormal data during data collection and transmission. Second, existing technologies mostly use centralized databases for data storage. Although this method is convenient for management and query, it is easily targeted by network attacks, and the security and reliability of data are difficult to guarantee. In addition, when existing systems process a large amount of data, the processing efficiency and real-time performance are not high. Especially when performing abnormal data location and decision support, the complexity of the algorithm and the large amount of data often lead to slow system response speed, affecting decision-making efficiency. Finally, existing technologies also have deficiencies in the traceability and immutability of abnormal data, and it is difficult to ensure the integrity and authenticity of data during storage and transmission. Summary of the Invention

[0004] In view of the above problems, the present invention is proposed.

[0005] Therefore, the technical problems solved by the present invention are: In the data collection process of existing power meter reading systems, there may be problems of data loss and inaccurate collection, and the security and reliability of data are difficult to guarantee. In the face of a large amount of meter reading data, existing systems have deficiencies in processing efficiency and real-time performance. Especially in abnormal data location and decision support, the complexity of the algorithm and the large amount of data often lead to slow system response speed.

[0006] To solve the above technical problems, the present invention provides the following technical solution: An abnormal location method based on meter reading data, including,

[0007] Collect multi-source data, use a linear regression model and a normalization function to fuse the collected multi-source data, and standardize the fused data to obtain a power system state matrix; by analyzing the power system state matrix, identify abnormal situations in the power system, quantify the degree of abnormality of each node, and based on the anomaly detection value model, determine whether the meter reading data is abnormal; according to historical data and current anomaly detection results, use a linear regression model and a normalization function to construct an intelligent decision-making model to evaluate the comprehensive decision-making score of the node, and based on the score result, determine the priority of the node and match the maintenance task; combine graph neural network to analyze the power grid topology structure, perform anomaly localization based on GPS and environmental data, determine the inaccessibility of the node, generate processing instructions, and encrypt and store the meter reading data.

[0008] As a preferred solution of the anomaly localization method based on meter reading data according to the present invention, wherein: the collection of multi-source data includes collecting meter reading data, historical data, weather data, geographical location data, time data, user type data, equipment status data, power load data and user behavior data through meters and sensors; the sensors include GPS, system clock and load monitoring devices.

[0009] As a preferred solution of the anomaly localization method based on meter reading data according to the present invention, wherein: the power system state matrix includes fusing all data sources collected at the same node through meters and sensors into one result, and converting the fused result into a power system state matrix through standardization processing;

[0010] The fusion processing includes fusing the collected multi-source data through a linear regression model and a normalization function to reflect the operating state of the power system and the specific conditions of each node, and the formula is expressed as:

[0011] ;

[0012] wherein, represents the result after fusing multi-source data, represents the total number of data sources, represents the th data source data, represents the th mean value of the data source data, represents the th standard deviation of the data source data, represents the integral variable, represents the th shape parameter of the gamma distribution of the data source data, represents the total number of data for normalization processing, represents the A normalized data, where i and j are variable indices.

[0013] As a preferred solution of the abnormal location method based on meter reading data according to the present invention, wherein: the power system state matrix further includes, according to the result of multi-source data fusion , perform data standardization processing using the Z-score formula to construct a power system state matrix, and the formula is expressed as:

[0014] ;

[0015] Wherein, represents the power system state matrix, represents the weight matrix, represents the input data matrix, represents the mean matrix, represents the identity matrix, represents the normalized data matrix, represents the parameter matrix;

[0016] The analysis of the power system state matrix includes reflecting the comprehensive parameters of the power system nodes by analyzing the power system state matrix, identifying abnormal situations in the power system, quantifying the abnormal degree of each node, and establishing an abnormal detection value model, and the formula is expressed as:

[0017] ;

[0018] Wherein, represents the abnormal detection value model, represents the total number of data sources, represents the fusion result of the th data source, is the state matrix of the th data source, and i is a variable index; The value range of is , when , it is determined that the meter reading data is normal; when

[0019] As a preferred solution of the abnormal location method based on meter reading data according to the present invention, wherein: the construction of the intelligent decision model includes constructing an intelligent decision model according to a linear regression model and a normalization function, and the formula is expressed as:

[0020] ;

[0021] ;

[0022] Among them, represents the comprehensive decision-making score of node ; represents the number of historical data records, represents the anomaly localization value of node in the l-th record, represents the maximum value of the anomaly localization values of all nodes, represents the inaccessibility value of node in the l-th record, represents the maximum value of the inaccessibility values of all nodes, represents the coefficient of the o-th feature in the l-th record, and M is the total number of features, represents the weight of the o-th feature, represents the o-th input feature, represents the intelligent decision support value of node ; represents the normalization constant, represents the length of the historical data of the node, represents the specific value of node in the r-th data record, represents the adjustment parameter of the r-th record, represents the integration variable, and l, v, o, and r are variable indices; The value range of is . When , the comprehensive decision-making weight of node is low; when , the comprehensive decision-making weight of node is medium; when , the comprehensive decision-making weight of node

[0023] The value range of is . If , then node is of low priority, it is judged that the node situation is stable, and the status of the node continues to be monitored; when , then node is of medium priority, it is judged that there are potential problems with the node, a maintenance task is generated, and the maintenance personnel will give priority to checking the node during the next inspection; when

[0024] is of high priority, a repair task is generated, and maintenance personnel are immediately dispatched to conduct a detailed inspection and repair of the node.As a preferred solution of the abnormal location method based on meter reading data according to the present invention, wherein: the intelligent decision-making further includes analyzing the power grid topology using a neural network and performing abnormal location based on GPS and environmental data, which is expressed by the formula:

[0025] ;

[0026] ;

[0027] After abnormal meter reading data is generated, combine GPS and environmental data to determine whether the current area is difficult to access, which is expressed by the formula:

[0028] ;

[0029] Wherein, represents the abnormal location value of node , G represents the number of layers of the graph neural network, represents the total number of nodes, represents the hidden state of node in the g-th layer of the graph neural network, and represent adjustment parameters, dist represents the geographical distance of node , represents the environmental data of node , is the activation function, represents the inaccessibility value of node , is the set of neighbor nodes of node v, is the normalization coefficient between node v and node u, is the weight matrix of neighbor node u in the g-th layer, is the hidden state of node u in the (g - 1)-th layer of the graph neural network, is the weight matrix of node v itself in the g-th layer, is the hidden state of node v in the (g - 1)-th layer, represents the GPS coordinate of node , represents the maximum value of GPS data, represents the maximum value of environmental data, is the abnormal detection value of node v, where g, v, and u are variable indices;

[0030] The value range of is , when , then node is normal; when There are slight anomalies, and an instruction for further monitoring is issued; when occurs, then the node has obvious anomalies, and an alarm message for immediate handling is issued;

[0031] The value range of is [0, 1]. When occurs, then the node is easily accessible, and an instruction for normal access is generated; when occurs, then the node is of medium difficulty to access, an inspection instruction is generated, the monitoring frequency is increased, and it is prompted that maintenance personnel wear safety equipment for maintenance; when occurs, then the node is difficult to access, an alarm is issued, and the problem is immediately determined through a remote diagnostic tool and maintenance is carried out.

[0032] As a preferred solution of the abnormal location method based on meter reading data described in the present invention, wherein: the encrypted storage of the meter reading data includes encrypting the meter reading data by using the AES symmetric encryption algorithm, encrypting the AES key by using the RSA asymmetric encryption algorithm, and storing the encrypted key in the key management system.

[0033] An abnormal location system based on meter reading data adopting the method described in any one of the present invention includes a data acquisition and processing module, an abnormal detection and analysis module, an intelligent decision-making and task assignment module, and an abnormal location and data security module.

[0034] The data acquisition and processing module is used to acquire multi-source data and perform fusion processing on the data through a linear regression model and a normalization function to obtain a power system state matrix.

[0035] The abnormal detection and analysis module includes an abnormal detection unit and an abnormal analysis unit, and is used to perform standardization processing on the fused data, construct a power system state matrix, identify abnormal conditions in the power system by analyzing the state matrix, quantify the abnormal degree of each node, and establish an abnormal detection value model.

[0036] The intelligent decision-making and task assignment module includes an intelligent decision-making unit and a task assignment unit, and is used to construct an intelligent decision-making model according to historical data and current abnormal detection results, evaluate the comprehensive decision-making score of the node, judge the priority of the node, and match maintenance tasks.

[0037] The abnormal location and data security module is used to analyze the power grid topology structure through a graph neural network, perform abnormal location in combination with GPS and environmental data, judge the inaccessibility of the node, generate processing instructions, and encrypt and store the meter reading data by using the AES symmetric encryption algorithm and the RSA asymmetric encryption algorithm.

[0038] A computer device, comprising: a memory and a processor; the memory stores a computer program, and when the processor executes the computer program, the steps of any one of the methods described in the present invention are implemented.

[0039] A computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of any one of the methods described in the present invention are implemented.

[0040] Advantages of the present invention: By collecting multi-source data, including electricity meter readings, environmental factors, etc., and using a linear regression model and a normalization function for fusion processing, the present invention realizes the comprehensive monitoring and in-depth understanding of the operation state of the power system; the fused data is processed by standardization to generate a power system state matrix that can reflect the real-time state of the power system, providing a reliable data basis for subsequent anomaly detection and analysis, and significantly improving the accuracy and efficiency of anomaly location; by analyzing the power system state matrix, abnormal situations can be discovered in a timely manner and the degree of abnormality of each node can be quantified. Based on the anomaly detection value model, the objectivity and systematicness of anomaly detection are improved, and the possibility of human misjudgment is reduced; by using historical data and current anomaly detection results, an intelligent decision-making model is constructed to evaluate the comprehensive decision-making score of nodes, and the node priorities are judged according to the score results, optimizing the allocation of maintenance resources and ensuring the efficiency and pertinence of maintenance work; by combining graph neural network analysis of the power grid topology structure, the present invention can more accurately locate anomalies and judge the inaccessibility of nodes, adjust the maintenance strategy, and improve the safety and efficiency of maintenance work; the meter reading data is encrypted and stored using the AES symmetric encryption algorithm and the RSA asymmetric encryption algorithm, ensuring data security, preventing data leakage, protecting the privacy of users and the system, and enhancing the reliability of the system and user trust. Description of the Drawings

[0041] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention, and those skilled in the art can obtain other drawings without creative efforts based on these drawings.

[0042] Figure 1 It is the overall flowchart of an anomaly location method based on meter reading data provided by the first embodiment of the present invention;

[0043] Figure 2 It is the system solution flowchart of an anomaly location system based on meter reading data provided by an embodiment of the present invention. Detailed Embodiments

[0044] To make the above objects, features, and advantages of the present invention more apparent and understandable, the following provides a detailed description of the specific embodiments of the present invention with reference to the accompanying drawings of the specification. Apparently, the described embodiments are only a part of the embodiments of the present invention, rather than all embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the scope of protection of the present invention.

[0045] Embodiment 1

[0046] Referring to Figure 1 , an embodiment of the present invention provides an abnormal location method based on meter reading data, including:

[0047] S1: Collect multi-source data, use a linear regression model and a normalization function to fuse the collected multi-source data, and standardize the fused data to obtain a power system state matrix.

[0048] The data collected by high-precision electricity meters and sensors includes meter reading data, historical data, weather data, geographical location data, time data, user type data, device status data, power load data, and user behavior data; the sensors include GPS, system clock, and load monitoring devices.

[0049] The power system state matrix includes fusing each data source collected at the same node by high-precision electricity meters and sensors into a comprehensive result, and standardizing the comprehensive result into a power system state matrix that can display the overall comprehensive operating state of each node;

[0050] The fusion process includes fusing the collected multi-source data through a linear regression model and a normalization function to reflect the operating state of the power system and the specific conditions of each node. The formula is expressed as:

[0051] ;

[0052] Among them, represents the result after fusing multi-source data, represents the total number of data sources, represents the th data source data, represents the mean of the th data source data, represents the th data source data standard deviation, represents the integral variable, represents the shape parameter of the gamma distribution of the th data source data, represents the total number of data after normalization processing, Indicates the th normalized data, where i and j are variable indices.

[0053] By fusing the collected multi-source data, the aim is to comprehensively utilize data from different sources to achieve comprehensive monitoring and analysis of the power system state; this comprehensiveness enables the system to better understand and predict power demand and load conditions. The fused data combines multi-dimensional information such as time, geographical location, and user behavior, etc., which helps in conducting more in-depth analysis; this multi-dimensional analysis can reveal complex relationships and patterns that cannot be reflected by a single data source. Fusing multi-source data can improve the robustness and reliability of the data. Even if one data source has an anomaly or error, the remaining data sources can still provide sufficient information for effective analysis and decision-making.

[0054] Normalizing the data can eliminate the dimensional differences between different data sources, enabling the information from different data sources to be compared and analyzed on the same platform, which helps improve the effect and accuracy of data fusion. Through the fusion of multi-source data, the system can more accurately monitor the operating state of the power system, identify potential problems and abnormal situations, which helps improve the safety and stability of the power system.

[0055] Furthermore, by comprehensively utilizing the analysis results of multi-source data, it can provide more scientific and reasonable decision-making support for the operation and dispatch of the power system. For example, it can predict changes in power demand, optimize the allocation of power resources, and improve the operating efficiency of the power system. Using the results of multi-source data fusion, an intelligent decision-making system can be constructed to further enhance the data processing and analysis capabilities of the power system. The intelligent processing can automatically identify and process abnormal data, improving the system's response speed and processing efficiency.

[0056] S2: By analyzing the power system state matrix, identify abnormal situations in the power system, quantify the degree of abnormality of each node, and based on the anomaly detection value model, determine whether the meter reading data is abnormal.

[0057] The power system state matrix also includes constructing the power system state matrix according to the results of multi-source data fusion and performing data standardization processing. The formula is expressed as:

[0058] ;

[0059] Where represents the power system state matrix, represents the weight matrix, represents the input data matrix, represents the mean matrix, represents the identity matrix, represents the normalized data matrix, Represents a parameter matrix;

[0060] The analyzed power system state matrix includes reflecting the comprehensive parameters of power system nodes through the analyzed power system state matrix, identifying abnormal situations in the power system, quantifying the degree of abnormality of each node, and establishing an abnormal detection value model, which is expressed by the formula:

[0061] ;

[0062] Among them, represents the abnormal detection value model, represents the total number of data sources, represents the fusion result of the th data source, is the state matrix of the th data source, and i is the variable index; When , it is determined that the meter reading data is normal; when , it is determined that there is a slight abnormality in the meter reading data, and a monitoring instruction is issued; when , it is determined that there is a serious abnormality in the meter reading data, and an alarm message for immediate processing is issued.

[0063] Establish a power system state matrix through the processed data, and use this state matrix to comprehensively analyze the power consumption system, aiming to monitor and evaluate the operating state of the power system in real time, detect and handle abnormal situations in a timely manner, and ensure the safety and stability of the power system. By fusing the data of multiple data sources, the state matrix can comprehensively reflect the operating state of the power system. The introduction of the weight matrix and the parameter matrix enables the contributions of different data sources to the system state to be weighed and adjusted.

[0064] By introducing the mean matrix and the identity matrix, the system can be dynamically adjusted according to the actual operating conditions to adapt to different power system environments. At the same time, the design of the state matrix has good scalability and can introduce new data sources and parameters as needed. The establishment of the state matrix enables the operating state of the power system to be monitored and evaluated in real time. By quantifying and modeling the states of each node, potential operating abnormalities and risks can be discovered in a timely manner. Comprehensive analysis of the power system state matrix can enhance the reliability and stability of the system. The system can make more accurate judgments and decisions based on real-time data to ensure the safe operation of the power system.

[0065] Furthermore, the abnormal data in the power system can be accurately identified through the anomaly detection formula. By using the ratio of the fusion result and the state matrix value, the degree of anomaly can be quantified, providing a basis for subsequent processing. The design of the anomaly detection value model can quantify the degree of anomaly, enabling hierarchical processing of anomalies. This design allows the system to take different countermeasures according to the severity of the anomaly, improving the pertinence and effectiveness of processing. By calculating and monitoring the anomaly detection value in real time, the system can achieve dynamic monitoring and early warning of abnormal situations. For minor anomalies, further monitoring can be carried out for confirmation; for obvious anomalies, processing measures can be taken immediately to prevent the problem from expanding; the anomaly detection model combines historical data and real-time data, providing strong support for the intelligent decision-making system. The system can make scientific and accurate decisions based on the anomaly detection results, improving the operation efficiency and management level of the power system. Through a data-driven approach, the system can continuously learn and optimize the anomaly detection algorithm, enhancing the accuracy and robustness of anomaly detection. This intelligent management method can better adapt to the complexity and dynamic changes of the power system, improving the system's adaptability.

[0066] S3: According to historical data and current anomaly detection results, use a linear regression model and a normalization function to construct an intelligent decision-making model, evaluate the comprehensive decision-making score of the node, and based on the score result, judge the priority of the node and match the maintenance task.

[0067] The construction of the intelligent decision-making model includes constructing an intelligent decision-making model according to a linear regression model and a normalization function, and the formula is expressed as:

[0068] ;

[0069] ;

[0070] Among them, represents the comprehensive decision-making score of node , represents the number of historical data records, represents the anomaly localization value of node in the l-th record, represents the maximum value of the anomaly localization values of all nodes, represents the inaccessibility value of node in the l-th record, represents the maximum value of the inaccessibility values of all nodes, represents the coefficient of the o-th feature in the l-th record, M is the total number of features, represents the weight of the o-th feature, represents the o-th input feature, represents the intelligent decision-making support value of node , represents the normalization constant, represents the length of the historical data of the node, represents the node specific value in the r-th data record, represents the adjustment parameter of the r-th record, represents the integration variable, where l, v, o, and r are variable indices; The value range of is If the comprehensive decision-making weight of the node is low;

[0071] If then the comprehensive decision-making weight of the node is medium; if then the comprehensive decision-making weight of the node is high;

[0072] The value range of is When the node is of low priority, it is judged that the node situation is stable, and the status of the node continues to be monitored; when When the node is of medium priority, it is judged that there are potential problems with the node, a maintenance task is generated, and the maintenance personnel will give priority to checking the node during the next inspection; when When the node is of high priority, a repair task is generated, and maintenance personnel are immediately dispatched to conduct a detailed inspection and repair of the node.

[0073] The intelligent decision-making also includes analyzing the power grid topology structure using a graph neural network and performing anomaly localization based on GPS and environmental data, and the formula is expressed as:

[0074] ;

[0075] ;

[0076] After the abnormal meter reading data is located, combined with GPS and environmental data, it is judged whether the current area is difficult to access, and the formula is expressed as:

[0077] ;

[0078] Among them, represents the anomaly localization value of the node G represents the number of layers of the graph neural network, represents the total number of nodes, represents the hidden state of the node in the g-th layer of the graph neural network, and represents the adjustment parameter, dist represents a node of the geographical distance represents a node of the environmental data is the activation function represents a node of the inaccessibility value is the set of neighbor nodes of node v is the normalization coefficient between node v and node u is the weight matrix of neighbor node u in the g-th layer is the hidden state of node u in the (g - 1)-th layer of the graph neural network is the weight matrix of node v itself in the g-th layer is the hidden state of node v in the (g - 1)-th layer represents a node of the GPS coordinates represents the maximum value of the GPS data represents the maximum value of the environmental data is the anomaly detection value of node v, where g, v, and u are variable indices;

[0079] The value range of is When then the node is normal; when then the node has a slight anomaly and issues an instruction for further monitoring; when then the node

[0080] The value range of is [0, 1]. When then the node is easily accessible and generates an instruction for normal access; when then the node is moderately difficult to access, generates an inspection instruction, increases the monitoring frequency, and prompts the maintenance personnel to wear safety equipment for maintenance; when is difficult to access, issues an alarm, and immediately determines the problem through a remote diagnostic tool and conducts maintenance.

[0081] Construct an intelligent decision-making model through a linear regression model and a normalization function to achieve a comprehensive decision-making score for each node in the power system, using historical data and real-time data to quantify the anomaly detection value and inaccessibility value of each node, thereby evaluating the comprehensive decision-making weight.

[0082] Combine the historical records and feature information of multiple data sources to ensure the comprehensiveness and accuracy of the decision-making model. Eliminate the differences in data dimensions through normalization processing to ensure that the information from different data sources can be compared and analyzed under the same standard. By quantifying the anomaly detection values and inaccessibility values of each node, the comprehensive decision-making weight of the node can be scientifically evaluated. The intelligent decision-making model can provide scientific and accurate decision-making support, optimizing resource allocation. By comprehensively analyzing the status and historical data of each node, improve the accuracy of anomaly detection and handling, and enhance the reliability and stability of the system.

[0083] Furthermore, use the anomaly analysis model and graph neural network (GNN) in combination for power grid topology analysis, and perform anomaly localization based on GPS and environmental data. Through the hierarchical structure of the neural network, fully explore the relationships between various nodes in the power grid, improve the accuracy of anomaly localization, and ensure timely handling. By comprehensively considering the power grid topology and multi-source data, improve the accuracy of anomaly detection and localization, and ensure the timely discovery and handling of problems. Use GNN and multi-source data to build an intelligent anomaly detection and localization system, improving the management level and response ability of the power system.

[0084] Furthermore, combine GPS and environmental data to evaluate the difficulty of accessing nodes and determine the action priorities of maintenance personnel. By quantifying the geographical location and environmental data of the nodes, evaluate whether the nodes are easily accessible. Classify and prioritize according to the difficulty of accessing nodes, which can improve maintenance efficiency and ensure that problems are solved in a timely manner. By scientifically evaluating the difficulty of accessing nodes, reasonably allocate maintenance resources, reducing unnecessary time and cost waste. By reasonably evaluating the difficulty of accessing nodes, optimize the maintenance plan and improve the maintenance management level of the power system. Perform maintenance and repair in a timely and accurate manner to ensure the stable operation of the power system and reduce the power outage risk caused by equipment failures.

[0085] S4: Analyze the power grid topology in combination with graph neural network, perform anomaly localization based on GPS and environmental data, judge the inaccessibility of nodes, generate processing instructions, and encrypt and store the meter reading data.

[0086] The encryption and storage of the meter reading data includes encrypting the meter reading data using the AES symmetric encryption algorithm, encrypting the AES key using the RSA asymmetric encryption algorithm, and storing the encrypted key in the key management system.

[0087] Furthermore, ensure the security of data access and prevent unauthorized access and data leakage. Multi-level verification and strict access control mechanisms can effectively prevent data from being illegally accessed and tampered with, ensuring the security and privacy of data. Through multi-level verification and access control, it can be ensured that only authorized users can access and decrypt the data, preventing data leakage and abuse, and enhancing the security and reliability of the system.

[0088] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not restrictive. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

[0089] Embodiment 2, referring to Figure 2 , is the second embodiment of the present invention. This embodiment provides an abnormal location system based on meter reading data, including a data collection and processing module, an abnormal detection and analysis module, an intelligent decision-making and task allocation module, and an abnormal location and data security module.

[0090] The data collection and processing module is used to collect multi-source data and perform fusion processing on the data through a linear regression model and a normalization function to obtain a power system state matrix.

[0091] The abnormal detection and analysis module includes an abnormal detection unit and an abnormal analysis unit, which are used to perform standardization processing on the fused data, construct a power system state matrix, identify abnormal situations in the power system by analyzing the state matrix, quantify the degree of abnormality of each node, and establish an abnormal detection value model.

[0092] The intelligent decision-making and task allocation module includes an intelligent decision-making unit and a task allocation unit, which are used to construct an intelligent decision-making model according to historical data and current abnormal detection results, evaluate the comprehensive decision-making score of the node, judge the priority of the node, and match maintenance tasks.

[0093] The abnormal location and data security module is used to analyze the power grid topology structure through a graph neural network, combine GPS and environmental data for abnormal location, judge the inaccessibility of the node, generate processing instructions, and encrypt and store the meter reading data using the AES symmetric encryption algorithm and the RSA asymmetric encryption algorithm.

[0094] Embodiment 3, the third embodiment of the present invention, which is different from the previous two embodiments in that:

[0095] If the above-described functions are implemented in the form of software function units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs.

[0096] The logic and / or steps represented in the flowchart or described in other ways herein, for example, can be considered as a predefined sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or in combination with these instruction execution systems, apparatus, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in combination with an instruction execution system, apparatus, or device.

[0097] More specific examples (non-exhaustive list) of computer-readable media include the following: an electrical connection portion with one or more wirings (electronic device), a portable computer diskette case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other media, then editing, interpreting, or otherwise processing it as appropriate, and then storing it in a computer memory.

[0098] It should be understood that each part of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one of the following techniques well known in the art or a combination thereof can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

Claims

1. A method for locating anomalies based on meter reading data, characterized in that: include: Collect multi-source data, use linear regression model and normalization function to fuse the collected multi-source data, and standardize the fused data to obtain the power system state matrix; By analyzing the power system state matrix, the abnormal conditions in the power system are identified, and the degree of abnormality of each node is quantified. Based on the abnormal detection value model, it is determined whether the meter reading data is abnormal; Based on historical data and current anomaly detection results, an intelligent decision-making model is constructed using a linear regression model and a normalization function to evaluate the comprehensive decision-making score of the node. Based on the score results, the node priority is determined and maintenance tasks are matched; Combined with graph neural network to analyze the grid topology, locate anomalies based on GPS and environmental data, determine the inaccessibility of nodes, generate processing instructions, and encrypt and store meter reading data; The constructing of the intelligent decision model includes constructing the intelligent decision model according to the linear regression model and the normalization function, and the formula is expressed as: Among them, P v represents the comprehensive decision score of node v, N represents the number of historical data records, and L vl represents the abnormal location value of node v in the lth record, max(L v ) represents the maximum value of all node abnormal location values, A vl represents the inaccessibility value of node v in the lth record, max(A v ) represents the maximum value of the inaccessibility value of all nodes, α lo represents the coefficient of the oth feature in the lth record, M is the total number of features, ω o represents the weight of the oth feature, z o , O v represents the intelligent decision support value of node v, Z represents the normalization constant, K represents the historical data length of the node, and R vr represents the specific value of node v in the rth data record, λ r represents the adjustment parameter of the rth record, t represents the integral variable, l, v, o and r are variable indexes; v The range of the value is [0,∞), when O v <0.1, the comprehensive decision weight of node v is low; when 0.1≤O v <0.3, the comprehensive decision weight of node v is medium; v When ≥0.3, the comprehensive decision weight of node v is high; P v The value range is [0,∞), when P v <0.1, the node v is of low priority, the node is judged to be stable, and the node status continues to be monitored; when 0.1≤P v <0.3, the node v is of medium priority, and it is judged that there is a potential problem with the node, and a maintenance task is generated. The maintenance personnel will give priority to checking the node in the next inspection. v When ≥0.3, the node v has high priority, a maintenance task is generated, and maintenance personnel are immediately dispatched to conduct a detailed inspection and repair of the node.

2. The method for locating anomalies based on meter reading data according to claim 1, characterized in that: The multi-source data collection includes collecting meter reading data, historical data, weather data, geographic location data, time data, user type data, equipment status data, power load data and user behavior data through electricity meters and sensors; the sensors include GPS, system clock and load monitoring device.

3. The abnormality locating method based on meter reading data according to claim 2, characterized in that: The power system state matrix includes fusing all data sources collected by electric meters and sensors to the same node into one result, and converting the fused result into the power system state matrix through standardized processing; The fusion processing includes fusing the collected multi-source data through a linear regression model and a normalization function to reflect the operation status of the power system and the specific conditions of each node. The formula is expressed as follows: Among them, F represents the result of multi-source data fusion, n represents the total number of data sources, and x i represents the data of the i-th data source, μ i represents the mean of the data from the i-th data source, σ i represents the standard deviation of the data from the ith data source, t represents the integral variable, k i represents the shape parameter of the gamma distribution of the i-th data source, m represents the total number of normalized data, and y j Represents the jth normalized data, where i and j are variable indexes.

4. The method for locating anomalies based on meter reading data according to claim 3, characterized in that: The power system state matrix also includes, according to the result F after multi-source data fusion, using the Z-score formula to perform data standardization processing, and constructing the power system state matrix, the formula is expressed as: S = W·(XM′)·(I+exp(Y·B)) Among them, S represents the power system state matrix, W represents the weight matrix, X represents the input data matrix, and M ' represents the mean matrix, I represents the identity matrix, Y represents the normalized data matrix, and B represents the parameter matrix; The analysis of the power system state matrix includes: identifying abnormal conditions in the power system by analyzing the comprehensive parameters of the power system nodes reflected by the power system state matrix, quantifying the abnormal degree of each node, and establishing an abnormal detection value model, which is expressed as follows: Where D represents the anomaly detection value model, n represents the total number of data sources, and F i represents the fusion result of the i-th data source, S i is the state matrix of the i data sources, i is the variable index; the value range of D is [0,∞), when D<0.1, the meter reading data is judged to be normal; when 0.1≤D<0.3, the meter reading data is judged to have a slight abnormality, and a monitoring instruction is issued; when D≥0.3, the meter reading data is judged to have a serious abnormality, and an alarm message for immediate processing is issued.

5. The method for locating anomalies based on meter reading data according to claim 4, characterized in that: The intelligent decision-making also includes using a neural network to analyze the topology of the power grid and locate abnormalities based on GPS and environmental data. The formula is expressed as: After locating the abnormal meter reading data, combine GPS and environmental data to determine whether the current area is difficult to access. The formula is: Among them, L v represents the abnormal location value of node v, G represents the number of layers of the graph neural network, V represents the total number of nodes, represents the hidden state of node v in the gth layer of the graph neural network, γ and represents the adjustment parameter, dist(v) represents the geographical distance of node v, and E v represents the environment data of node v, ρ is the activation function, A v represents the inaccessibility value of node v, is the set of neighbor nodes of node v, c vu is the normalized coefficient between node v and node u, is the weight matrix of the neighbor node u in the gth layer, is the hidden state of node u in the g-1th layer of the graph neural network, is the weight matrix of the g-th layer node v itself, is the hidden state of node v in the g-1th layer, GPS v represents the GPS coordinates of node v, max(GPS) represents the maximum value of GPS data, max(E) represents the maximum value of environmental data, D v is the abnormal detection value of node v, g, v and u are variable indexes; L v The range of L is [0,∞). v <0.1, the node v is normal; when 0.1≤L v <0.3, there is a slight abnormality in node v, and a further monitoring instruction is issued; when L v When ≥0.3, there is an obvious abnormality in node v, and an alarm message for immediate processing is issued; A v The value range is [0, 1], when A v <0.3, the node v is easily accessible and generates a normal approach instruction; when 0.3≤A v <0.7, the node v is approached with medium difficulty, and an inspection instruction is generated to increase the monitoring frequency and remind the maintenance personnel to wear safety equipment for maintenance. v When ≥0.7, the node v is difficult to access, an alarm is issued, and the problem is immediately identified through remote diagnostic tools and maintenance is performed.

6. The abnormality locating method based on meter reading data according to claim 5, characterized in that: The encrypted storage of the meter reading data includes: encrypting the meter reading data using an AES symmetric encryption algorithm, encrypting the AES key using an RSA asymmetric encryption algorithm, and storing the encrypted key in a key management system.

7. A system using the abnormality locating method based on meter reading data as claimed in any one of claims 1 to 6, characterized in that: It includes data collection and processing module, anomaly detection and analysis module, intelligent decision-making and task allocation module, and anomaly location and data security module; The data acquisition and processing module is used to collect multi-source data and perform data fusion processing through a linear regression model and a normalization function to obtain a power system state matrix; The anomaly detection and analysis module includes an anomaly detection unit and an anomaly analysis unit, which are used to standardize the fused data, construct a power system state matrix, identify abnormal conditions in the power system by analyzing the state matrix, quantify the degree of abnormality of each node, and establish an anomaly detection value model; The intelligent decision-making and task allocation module includes an intelligent decision-making unit and a task allocation unit, which are used to build an intelligent decision-making model based on historical data and current anomaly detection results, evaluate the comprehensive decision score of the node, determine the priority of the node, and match the maintenance task; The anomaly location and data security module is used to analyze the power grid topology through a graph neural network, and to locate anomalies in combination with GPS and environmental data, determine the inaccessibility of nodes, and generate processing instructions, and use the AES symmetric encryption algorithm and the RSA asymmetric encryption algorithm to encrypt and store meter reading data.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the abnormality locating method based on meter reading data described in any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the abnormality locating method based on meter reading data according to any one of claims 1 to 6 are implemented.

Images