Intelligent management electric energy metering box with multiple safety protection and metering method

Abstract

The invention provides an intelligent management electric energy metering box with multiple safety protection and a metering method. Belongs to the technical field of intelligent power grid and electricity utilization information acquisition. The method comprises the following steps: based on a physical structure of a new standard metering box of a state grid, carrying out two-channel electricity larceny prevention sensing network deployment, and generating electricity larceny prevention sensing layout data; according to the electricity larceny prevention sensing layout data, a current anomaly detection sensor and a magnetic field disturbance monitoring sensor are integrated, and a two-channel electricity larceny behavior sensing network is constructed; by integrating dual-channel electricity larceny prevention sensing, multi-parameter edge monitoring, lightweight AI diagnosis, block chain evidence storage and active operation and maintenance strategies, the safety and reliability of electric energy metering are significantly improved.

Description

Technical Field

[0001] This invention proposes an intelligent management power metering box and metering method with multiple safety protections, belonging to the field of smart grid and power consumption information collection technology. Background Technology

[0002] With the rapid development of smart grids, electricity metering, as a core component of electricity trading and management, is directly related to the stable operation of the power grid and the legitimate rights and interests of users.

[0003] However, traditional electricity metering boxes, even those designed according to the new State Grid standards, such as the HW-SX / PX series, still lack comprehensive protection, relying mainly on mechanical seals and simple circuit breakers. This makes them ill-equipped to effectively detect covert methods such as meter tampering and current diversion for electricity theft. Furthermore, status monitoring is significantly delayed; detection of electrical hazards such as temperature rise, partial discharge, and arcing depends on manual inspections, failing to provide real-time warnings and increasing maintenance risks and costs.

[0004] In addition, data security is also a weak point. The collected data is mostly transmitted in plaintext, which makes it extremely vulnerable to man-in-the-middle attacks or local tampering, threatening the authenticity and integrity of the data.

[0005] In terms of operation and maintenance, traditional methods are mostly passive, relying on user reports for repairs. It is difficult to predict gradual faults such as aging switches and loose connectors, resulting in untimely fault handling and affecting power supply quality.

[0006] In view of this, there is an urgent need to make breakthroughs in existing technologies and build a new power metering method that integrates physical security, electrical protection, intelligent monitoring and trusted computing, so as to realize a five-dimensional protection system of "end-edge-cloud" collaboration and improve the security, reliability and operation and maintenance efficiency of power metering. Summary of the Invention

[0007] This invention provides an intelligent energy metering box with multiple safety protections and a metering method to solve the problems mentioned in the background art above:

[0008] This invention proposes a smart energy metering method with multiple safety protections, the method comprising:

[0009] S1. Based on the physical structure of the State Grid's new standard metering box, deploy a dual-channel anti-theft sensor network to generate anti-theft sensor layout data; based on the anti-theft sensor layout data, integrate current anomaly detection sensors and magnetic field disturbance monitoring sensors to construct a dual-channel electricity theft behavior perception network.

[0010] S2. Based on the dual-channel electricity theft behavior perception network, configure multi-parameter edge monitoring nodes, collect multi-source data, and generate multi-source electrical parameter monitoring data; perform edge-end preprocessing on the multi-source electrical parameter monitoring data to generate structured edge monitoring data;

[0011] S3. Based on structured edge monitoring data, load a lightweight AI diagnostic model to identify electricity theft behavior patterns and diagnose electrical hazards in real time, generating electricity theft risk warning data and equipment health status assessment data; classify the electricity theft risk warning data and equipment health status assessment data into risk levels and generate a four-level warning instruction set;

[0012] S4. Encrypt and store the four-level early warning instruction set through blockchain nodes to generate an immutable early warning event chain; simultaneously upload structured edge monitoring data to the cloud management platform to build a collaborative data flow between the end, edge, and cloud.

[0013] S5. Based on the analysis results of the early warning event chain and the cloud management platform, execute proactive operation and maintenance strategies and generate dynamic operation and maintenance task work orders;

[0014] S6. Based on the operation and maintenance effect data fed back by dynamic operation and maintenance task work orders, iteratively optimize the parameters of the lightweight AI diagnostic model and the blockchain evidence storage strategy to form an adaptive security protection closed-loop system, continuously enhancing the security defense capability and operation and maintenance efficiency of the power metering device.

[0015] This invention proposes a metering box for implementing the intelligent management power metering method with multiple safety protections as described above, the metering box comprising:

[0016] Network Deployment Module: Based on the physical structure of the State Grid's new standard metering box, a dual-channel anti-theft sensor network is deployed to generate anti-theft sensor layout data; based on the anti-theft sensor layout data, current anomaly detection sensors and magnetic field disturbance monitoring sensors are integrated to construct a dual-channel electricity theft behavior perception network;

[0017] Data acquisition module: Based on the dual-channel electricity theft behavior perception network, configure multi-parameter edge monitoring nodes, collect multi-source data, and generate multi-source electrical parameter monitoring data; perform edge-end preprocessing on the multi-source electrical parameter monitoring data to generate structured edge monitoring data;

[0018] Level Classification Module: Based on structured edge monitoring data, a lightweight AI diagnostic model is loaded to identify electricity theft behavior patterns and diagnose electrical hazards in real time, generating electricity theft risk warning data and equipment health status assessment data; the electricity theft risk warning data and equipment health status assessment data are classified into risk levels, and a four-level warning instruction set is generated;

[0019] Platform management module: It encrypts and stores the four-level early warning instruction set through blockchain nodes to generate an immutable early warning event chain; it also uploads structured edge monitoring data to the cloud management platform to build a collaborative data flow between the end, edge and cloud.

[0020] Dynamic Operation and Maintenance Module: Based on the analysis results of the early warning event chain and the cloud management platform, it executes proactive operation and maintenance strategies and generates dynamic operation and maintenance task work orders;

[0021] Security Defense Module: Based on the operation and maintenance effect data fed back by dynamic operation and maintenance task work orders, the module iteratively optimizes the parameters of the lightweight AI diagnostic model and the blockchain evidence storage strategy to form an adaptive security protection closed-loop system, continuously enhancing the security defense capabilities and operation and maintenance efficiency of the power metering device.

[0022] The beneficial effects of this invention are as follows: By integrating dual-channel anti-theft sensing, multi-parameter edge monitoring, lightweight AI diagnostics, blockchain evidence storage, and proactive maintenance strategies, the security and reliability of electricity metering are significantly improved. This method reduces electricity loss caused by covert theft, enhances real-time perception and early warning capabilities for electrical hazards, and reduces the frequency and cost of manual inspections. Simultaneously, blockchain evidence storage technology avoids the risk of data tampering during transmission, ensuring data authenticity and integrity. Furthermore, the implementation of proactive maintenance strategies can both predict and handle gradual faults such as switch aging and loose connectors, and quickly respond to theft risk warnings, effectively preventing power outages caused by untimely fault handling. Overall, this method constructs a five-dimensional protection system integrating end, edge, and cloud, providing a highly secure, highly reliable, and auditable next-generation electricity metering solution for smart grids. Attached Figure Description

[0023] Figure 1 This is a diagram illustrating the steps of the method described in this invention;

[0024] Figure 2 This is a schematic diagram of the metering box module described in this invention. Detailed Implementation

[0025] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0026] One embodiment of the present invention, such as Figure 1 As shown, a smart energy metering method with multiple safety protections is disclosed, the method comprising:

[0027] S1. Based on the physical structure of the State Grid's new standard metering box, deploy a dual-channel anti-theft sensor network to generate anti-theft sensor layout data; based on the anti-theft sensor layout data, integrate current anomaly detection sensors and magnetic field disturbance monitoring sensors to construct a dual-channel electricity theft behavior perception network.

[0028] S2. Based on the dual-channel electricity theft behavior perception network, configure multi-parameter edge monitoring nodes, collect multi-source data, including current, voltage, temperature, partial discharge and arc characteristic signals, and generate multi-source electrical parameter monitoring data; perform edge preprocessing on the multi-source electrical parameter monitoring data, including signal filtering, feature extraction and time sequence alignment, to generate structured edge monitoring data;

[0029] S3. Based on structured edge monitoring data, load a lightweight AI diagnostic model to identify electricity theft behavior patterns and diagnose electrical hazards in real time, generating electricity theft risk warning data and equipment health status assessment data; classify the electricity theft risk warning data and equipment health status assessment data into risk levels and generate a four-level warning instruction set;

[0030] S4. Encrypt and store the four-level early warning instruction set through blockchain nodes to generate an immutable early warning event chain; simultaneously upload structured edge monitoring data to the cloud management platform to build a collaborative data flow between the end, edge, and cloud.

[0031] S5. Based on the analysis results of the early warning event chain and the cloud management platform, execute proactive operation and maintenance strategies, including remote switch opening and closing control, connector loosening early warning and switch aging life prediction, and generate dynamic operation and maintenance task work orders.

[0032] S6. Based on the operation and maintenance effect data fed back by dynamic operation and maintenance task work orders, iteratively optimize the parameters of the lightweight AI diagnostic model and the blockchain evidence storage strategy to form an adaptive security protection closed-loop system, continuously enhancing the security defense capability and operation and maintenance efficiency of the power metering device.

[0033] The working principle and effects of the above technical solution are as follows: Through a multi-layered intelligent management method for electricity metering with multiple security protections, it significantly improves the accuracy of identifying electricity theft and electrical hazards, reducing missed and false diagnoses. It enhances the timeliness of electrical safety early warnings, preventing the escalation of power risks caused by equipment failures. It improves the security of data storage and transmission, preventing information tampering and making maintenance decisions more credible. It reduces the intensity and cost of manual maintenance, minimizing ineffective resource investment. Edge-cloud collaboration accelerates data flow efficiency, and closed-loop optimization continuously strengthens security defense capabilities. It ensures accurate electricity metering while improving maintenance response speed, avoiding economic losses caused by insufficient protection or delayed maintenance, making electricity metering management more efficient and reliable.

[0034] In one embodiment of the present invention, S1 includes:

[0035] S11. Analyze the spatial layout of the SMC / PC+ABS non-metallic enclosure, clarify the installation position and distribution pattern of the core components, including the front switch, the rear switch and connectors; combine the differences in the number of meters in the metering box (single-phase one meter to nine meters, three-phase one meter to four meters, etc.), delineate the key coverage area of ​​the sensor monitoring, and generate anti-theft sensor layout data.

[0036] S12. Select current anomaly detection sensors and magnetic field disturbance monitoring sensors that are compatible with the rated current range of different metering boxes (100A, 63A, 80A, etc.); plan the installation spacing and fixing method of the sensors, disconnect switches, and miniature circuit breakers to ensure that the original electrical connection and heat dissipation of the metering box are not affected.

[0037] S13. Perform parameter calibration on the two sensors to match the detection accuracy with the operating parameter range of the metering box and eliminate the error caused by electromagnetic interference from the non-metallic box.

[0038] S14. Integrate and install the calibrated sensors according to the layout data to form a dual-channel electricity theft detection network covering the key areas of the enclosure.

[0039] The working principle and effects of the above technical solution are as follows: By analyzing the layout of the non-metallic enclosure and delineating key monitoring areas, the accuracy of sensor coverage is improved, avoiding monitoring blind spots. Sensors with different rated currents are selected and reasonable installation spacing and fixing methods are planned to enhance the compatibility between sensors and the metering box, reducing the impact of improper installation on existing electrical connections and heat dissipation. Electromagnetic interference errors are eliminated through parameter calibration, improving the accuracy of detection data and reducing false alarms and missed alarms. An integrated dual-channel sensing network is formed, capable of capturing abnormal current changes and monitoring magnetic field disturbances, enhancing the comprehensiveness of anti-theft monitoring, avoiding missed detections of electricity theft due to a single monitoring mode, thereby improving the security level of electricity metering, reducing economic losses caused by electricity theft, and making metering monitoring more stable and reliable.

[0040] In one embodiment of the present invention, S2 includes:

[0041] S21. For key components inside the metering box, including disconnect switches, current transformers, and connectors, deploy multi-parameter edge monitoring nodes to ensure that each core component corresponds to an independent monitoring unit; synchronously collect current, voltage, temperature, partial discharge, and arc characteristic signals through the edge monitoring nodes, and summarize them to form multi-source electrical parameter monitoring data;

[0042] S22. Use a digital filtering algorithm to process the collected data and remove invalid information caused by box vibration and environmental electromagnetic interference;

[0043] S23. Extract the core feature quantities related to electricity theft (abnormal current fluctuations) and equipment failure (excessive temperature, partial discharge signal) from the data;

[0044] S24. Time-axis alignment is performed on the time-series data of different monitoring nodes to unify the data acquisition frequency and format standards; the filtered data and core feature quantities are integrated to generate structured edge monitoring data, which has a high degree of structure and can be directly used for analysis.

[0045] The working principle and effects of the above technical solution are as follows: Independent monitoring nodes are deployed for each core component, simultaneously collecting multiple electrical parameters to improve the comprehensiveness of data acquisition and avoid missing key operational information. Digital filtering effectively removes invalid data caused by environmental interference and enclosure vibration, reducing the impact of noise on analysis results and improving data purity. Core feature quantities related to electricity theft and faults are extracted, making the data more valuable for analysis and enhancing the targeting of subsequent identification and judgment. Time alignment and format unification form a standardized system from scattered data, reducing the processing difficulty of subsequent AI diagnosis and avoiding analysis delays or misjudgments caused by data chaos. This provides high-quality data support for subsequent risk warnings, accelerates data processing efficiency, makes the investigation of electrical hazards and electricity theft more accurate and efficient, reduces judgment bias caused by data problems, and ensures the reliability of metering management.

[0046] In one embodiment of the present invention, S24 includes:

[0047] Extract the original timestamps of the time series data from each monitoring node to generate independent time series datasets for each node; based on the time series datasets of each node, establish a unified time reference scale to eliminate the time difference in data collection between different nodes.

[0048] According to a unified time base scale, interpolation is performed on the time series data of each node to fill the acquisition gaps and form a continuous data chain.

[0049] Unify the data acquisition frequency of all nodes to ensure consistent update intervals for various types of time-series data; standardize the data format of different types of parameters, including the expression form, such as current, voltage, and temperature.

[0050] By integrating the aligned, completed, and standardized filtered data with core feature quantities, structured edge detection data is generated.

[0051] The working principle and effects of the above technical solution are as follows: Extracting the original timestamps of each node and establishing a unified benchmark eliminates time differences in data collection, avoiding analytical biases caused by data asynchrony. Interpolation fills in the gaps in data collection, allowing data to form a continuous chain and reducing the impact of breakpoints on judgment. Unifying the collection frequency and data format improves the consistency of various parameters and reduces the complexity of subsequent processing. The integrated structured data enhances the usability of direct analysis, enabling rapid adaptation to AI diagnostic models and shortening the data flow cycle, avoiding delays in hazard investigation due to data clutter, and making data analysis of electricity metering more efficient and accurate.

[0052] In one embodiment of the present invention, S3 includes:

[0053] S31. Load a lightweight AI diagnostic model adapted to the metering box's operating scenario, and import historical operating data and fault cases of different models of metering boxes;

[0054] S32. The model uses deep learning to understand the electrical parameter change patterns corresponding to different electricity theft behaviors (such as wiring tampering and magnetic field interference); it also learns the characteristic manifestations of common hidden dangers, including aging of miniature circuit breakers and poor contact of connectors, to improve the diagnostic logic.

[0055] S33. Input the structured edge monitoring data into the model to generate electricity theft risk warning data and equipment health status assessment data; refer to the State Grid safety standards and the rated parameters of the metering box to set four-level risk assessment thresholds (minor, moderate, relatively severe, and serious).

[0056] S34. Based on the threshold, the early warning data and the evaluation data are classified and judged to generate a four-level early warning instruction set, which includes the handling priority.

[0057] The working principle and effects of the above technical solution are as follows: A lightweight AI diagnostic model adapted to the metering box scenario is loaded, and historical operating data and fault cases from different models are imported, allowing the model to accurately grasp the characteristics and patterns of electricity theft and equipment hazards, improving the targeting and accuracy of diagnosis. Simultaneously learning the electrical parameter changes of both types of problems enhances the comprehensiveness of risk identification and reduces missed diagnoses caused by single-dimensional diagnosis. Structured edge monitoring data is input into the model to generate early warning and assessment data. Four-level thresholds are set based on State Grid safety standards and metering box rated parameters, making risk classification more aligned with actual operational needs and avoiding misjudgments caused by threshold imbalances. An instruction set containing handling priorities is generated based on the thresholds, which not only allows maintenance work to proceed in an orderly manner, reducing ineffective resource investment, but also accelerates the response speed to high-risk problems, preventing the expansion of hazards due to chaotic handling sequences, making risk management of electricity metering more efficient and reliable.

[0058] In one embodiment of the present invention, S34 includes:

[0059] Organize electricity theft risk warning data and equipment health status assessment data to form a data set to be judged;

[0060] Each data point in the dataset to be judged is compared with the four-level risk assessment threshold one by one, and the matching risk level is marked.

[0061] Based on the importance of key components within the metering box, including disconnect switches, connectors, and miniature circuit breakers, different risk levels are assigned to different handling priorities; the risk level of each data item is associated with its corresponding handling priority to form preliminary early warning instruction items;

[0062] Verify whether the initial instruction entries conform to the State Grid safety standards and the metering box's rated parameters, and correct any mismatches.

[0063] All verified early warning instruction entries are integrated, sorted by risk level and handling priority, and a four-level early warning instruction set is generated.

[0064] The working principle and effects of the above technical solution are as follows: Data is organized to form a set of issues to be assessed; each item is compared against thresholds to indicate its risk level, improving the accuracy of risk assessment and reducing misjudgments and omissions. Prioritization of handling is based on the importance of key components such as disconnect switches and connectors, allowing maintenance resources to be precisely targeted at core issues, improving response efficiency, and preventing delays in handling high-risk hazards due to indiscriminate use. Verification of instruction entries ensures they conform to State Grid safety standards and metering box rated parameters, correcting mismatches and enhancing instruction compliance to prevent electrical safety issues caused by improper operations. The integrated and sorted early warning instruction set is both clear and easy for frontline execution, ensuring that high-risk issues are addressed first, reducing internal maintenance friction caused by chaotic instructions, and making risk management for electricity metering more orderly and targeted.

[0065] In one embodiment of the present invention, step S4 includes:

[0066] S41. Distribute the four-level early warning instruction set to the distributed blockchain nodes in timestamp order to ensure that each instruction corresponds to an independent evidence storage unit; the blockchain nodes use an asymmetric encryption algorithm to encrypt the instruction set and generate an unalterable encrypted data packet.

[0067] S42. Connect the encrypted data of each node to form a time-continuous and traceable early warning event chain, and record the early warning time, level and related monitoring data;

[0068] S43. Extract the core information directly related to measurement accuracy and equipment status from the structured edge monitoring data, and remove redundant data;

[0069] S44. According to the preset transmission protocol and data format, the core information is synchronously uploaded to the cloud management platform to ensure the stability of data transmission;

[0070] S45. Establish a real-time communication link between edge monitoring nodes and the cloud management platform, integrate the data flow path of end-side acquisition, edge-side processing, and cloud analysis, and construct a collaborative data flow between the end, edge, and cloud.

[0071] The working principle and effects of the above technical solution are as follows: Blockchain encryption enhances the security of early warning commands, preventing data tampering; independent evidence storage units ensure traceability of each command, reducing subsequent disputes over liability. Connecting encrypted data from each node forms an event chain, fully recording key early warning information, improving the convenience of problem backtracking, and preventing the omission of crucial details. Eliminating redundant data reduces transmission and storage pressure, allowing core information to be more focused and improving data processing efficiency. Standardized transmission protocols and formats ensure stable data uploads, avoiding transmission interruptions or data distortion. Edge-cloud collaborative data flow connects the entire path of collection, processing, and analysis, enabling rapid information flow and allowing cloud-based decision-making to better align with actual on-site conditions, avoiding response delays caused by data silos, and overall improving the reliability and collaborative efficiency of electricity metering data management.

[0072] In one embodiment of the present invention, S41 includes:

[0073] S411. Organize the four-level early warning instruction set, extract the generation time information of each instruction, and sort them in chronological order according to timestamps; assign a unique identifier code to each sorted instruction to ensure that each evidence storage unit can be traced individually.

[0074] S412. Analyze the current load status of each node in the distributed blockchain and distribute the sorted instruction set according to the load balancing principle.

[0075] S413. Each blockchain node receives the corresponding instruction, verifies the instruction identifier code and content integrity, and excludes damaged data.

[0076] S414. The node initiates an asymmetric encryption algorithm to encrypt the verified instructions and convert them into ciphertext. It then aggregates the encrypted individual instructions from each node and integrates them to form a unified and tamper-proof encrypted data packet.

[0077] The working principle and effects of the above technical solution are as follows: Sorting by timestamp and assigning a unique identifier to each instruction improves the traceability convenience of a single evidence storage unit, avoiding the problem of inaccurate instruction location during subsequent verification. Analyzing the load status of each blockchain node and distributing instructions evenly reduces the operating pressure on individual nodes, avoiding processing delays or lags caused by overload of some nodes. Verifying the integrity of instruction identifiers and content eliminates damaged data, reduces interference from invalid information in the evidence storage process, and enhances the reliability of data transmission. Converting instructions into ciphertext and integrating data packets through asymmetric encryption enhances the security protection capabilities of warning instructions, preventing data from being tampered with or stolen during the evidence storage process. This ensures both the orderly storage of instructions and improves the efficiency and security of evidence storage, making the evidence storage of early warning information for electricity metering more standardized and credible, providing solid support for subsequent operation and maintenance traceability.

[0078] In one embodiment of the present invention, S412 includes:

[0079] Collect real-time operational data from each node of the distributed blockchain, including CPU utilization, storage usage, and data processing queue length, and generate a node load status dataset.

[0080] Based on the load status dataset, calculate the current load pressure value of each node and set the load balancing baseline threshold.

[0081] By comparing the load pressure values ​​of each node with the baseline threshold, the nodes are classified into three levels: light load, medium load, and heavy load.

[0082] The data volume and processing complexity of a single instruction in the sorted instruction set are statistically analyzed, and the load requirements of each instruction are marked.

[0083] Match instruction load requirements with node levels, and formulate distribution rules to allocate heavy-load instructions to light-load nodes and medium-load instructions to medium-load nodes;

[0084] Plan the data transmission path between nodes, avoid overloaded node transmission channels, and accurately distribute the sorted instruction set to the corresponding blockchain nodes according to the rules.

[0085] The working principle and effects of the above technical solution are as follows: It collects real-time operational data from each blockchain node to generate a load dataset, making the node's operational status clearly visible and avoiding resource waste caused by blindly distributing instructions. It calculates load pressure values ​​and sets benchmark thresholds, classifying nodes into three levels to improve the accuracy of load assessment and reduce distribution imbalances. It labels the load requirements of each instruction and matches it to nodes of the corresponding level, reducing the risk of lightly loaded nodes being idle and heavily loaded nodes being overloaded, thus enhancing resource utilization efficiency. It plans transmission paths to avoid heavily loaded channels, improving the smoothness of instruction distribution and avoiding processing delays caused by transmission congestion. This not only makes the overall operation of blockchain nodes more stable but also ensures the timeliness of early warning instruction processing, reducing system lag or data backlog caused by uneven load, making the encrypted evidence storage process more efficient and reliable, and providing stable support for subsequent data traceability.

[0086] In one embodiment of the present invention, step S5 includes:

[0087] S51, the cloud management platform receives early warning event chains and structured edge monitoring data, and performs comprehensive analysis by combining the metering box's historical operation and maintenance records with model characteristics;

[0088] S52. For serious or higher levels of electricity theft risk, automatically trigger remote control signals to precisely control the opening and closing of the isolating switch and cut off the illegal power circuit.

[0089] S53. Potential hazards in the docking equipment, including loose plugs and abnormal temperature of miniature circuit breakers, generate early warning prompts, the early warning prompts include specific locations and risk descriptions; based on transformer operating data, switch operation frequency and temperature change trends, predict the remaining lifespan of the equipment and mark the parts that need to be replaced in advance.

[0090] S54. Based on the risk level, equipment importance, and maintenance resources, clarify the execution priority and operation process of each maintenance task;

[0091] S55. Integrate remote control commands, hidden danger warning prompts, life prediction results and maintenance priorities to generate dynamic maintenance task work orders. The dynamic maintenance task work orders are well-organized and can be executed directly.

[0092] The working principle and effects of the above technical solution are as follows: Comprehensive analysis of the early warning event chain, monitoring data, historical maintenance records, and model characteristics allows maintenance decisions to be more aligned with the actual operating conditions of the metering box, avoiding improper handling due to biased judgments. For high-level electricity theft risks, remote control of isolating switches quickly cuts off illegal circuits, improving the timeliness of electricity theft response and reducing economic losses from power loss. Accurate labeling of equipment hazard locations and risks, combined with operational data to predict component aging lifespan, identifies replacement components in advance, preventing escalation of faults and potential electrical safety issues. Prioritizing maintenance based on risk level and resource availability, and clarifying operating procedures, ensures more rational resource allocation and reduces ineffective investment and process chaos. Dynamic work orders generated by integrating various information are clearly structured, allowing maintenance personnel to execute them directly while reducing the probability of operational errors, thus enhancing the overall targeting and efficiency of electricity metering maintenance, making the electricity environment safer and maintenance management more worry-free.

[0093] In one embodiment of the present invention, step S6 includes:

[0094] S61. Collect execution effect data from dynamic operation and maintenance task work orders, including switch control response speed, equipment parameter recovery status after hazard handling, and electricity theft behavior handling results; compare monitoring data before and after operation and maintenance, and analyze the deviation rate, missed judgment rate, and false judgment rate of the lightweight AI diagnostic model in electricity theft identification and hazard judgment.

[0095] S62. Based on the analysis results, adjust the internal algorithm parameters of the model, optimize the feature recognition weights of electricity theft and equipment hazards, and improve diagnostic accuracy; optimize the encryption algorithm and data synchronization frequency of blockchain nodes, balance data security and transmission efficiency, and enhance the integrity of the early warning event chain.

[0096] S63. Based on the actual operation and maintenance results, adjust the risk assessment threshold and early warning classification standards to make the early warning instructions more in line with the on-site operation and maintenance needs; integrate the model parameter optimization results, blockchain evidence storage strategy adjustment scheme and threshold revision standards to form an adaptive security protection closed-loop system, and continuously enhance the security defense capabilities and operation and maintenance efficiency of the power metering device.

[0097] The working principle and effects of the above technical solution are as follows: Collecting and comparing operational data accurately identifies deviations, omissions, and misjudgments in the AI ​​diagnostic model, providing a practical basis for parameter adjustment, improving model diagnostic accuracy, and reducing subsequent judgment errors. Adjusting model algorithm parameters and feature recognition weights makes the identification of electricity theft and equipment hazards more accurate; optimizing blockchain encryption algorithms and synchronization frequencies balances data security and transmission efficiency, enhances the integrity of the early warning event chain, and avoids vulnerabilities in data storage or transmission. Revising risk assessment thresholds and grading standards based on actual operational results makes early warning instructions more aligned with on-site needs, avoiding operational chaos caused by unreasonable thresholds. Integrating various optimization solutions to form an adaptive closed-loop system continuously strengthens the security defense capabilities of electricity metering devices, continuously improves operational efficiency, reduces ineffective resource investment, prevents system rigidity from failing to adapt to actual operational changes, and ensures that electricity metering management remains highly efficient and reliable.

[0098] According to one embodiment of the present invention, a metering box for implementing the intelligent management power metering method with multiple safety protections as described above is provided, the metering box comprising:

[0099] Network Deployment Module: Based on the physical structure of the State Grid's new standard metering box, a dual-channel anti-theft sensor network is deployed to generate anti-theft sensor layout data; based on the anti-theft sensor layout data, current anomaly detection sensors and magnetic field disturbance monitoring sensors are integrated to construct a dual-channel electricity theft behavior perception network;

[0100] Data acquisition module: Based on the dual-channel electricity theft behavior perception network, configure multi-parameter edge monitoring nodes, collect multi-source data, including current, voltage, temperature, partial discharge and arc characteristic signals, and generate multi-source electrical parameter monitoring data; perform edge-end preprocessing on the multi-source electrical parameter monitoring data, including signal filtering, feature extraction and time sequence alignment, to generate structured edge monitoring data;

[0101] Level Classification Module: Based on structured edge monitoring data, a lightweight AI diagnostic model is loaded to identify electricity theft behavior patterns and diagnose electrical hazards in real time, generating electricity theft risk warning data and equipment health status assessment data; the electricity theft risk warning data and equipment health status assessment data are classified into risk levels, and a four-level warning instruction set is generated;

[0102] Platform management module: It encrypts and stores the four-level early warning instruction set through blockchain nodes to generate an immutable early warning event chain; it also uploads structured edge monitoring data to the cloud management platform to build a collaborative data flow between the end, edge and cloud.

[0103] Dynamic Operation and Maintenance Module: Based on the analysis results of the early warning event chain and the cloud management platform, it executes proactive operation and maintenance strategies, including remote switch opening and closing control, connector loosening early warning and switch aging life prediction, and generates dynamic operation and maintenance task work orders;

[0104] Security Defense Module: Based on the operation and maintenance effect data fed back by dynamic operation and maintenance task work orders, the module iteratively optimizes the parameters of the lightweight AI diagnostic model and the blockchain evidence storage strategy to form an adaptive security protection closed-loop system, continuously enhancing the security defense capabilities and operation and maintenance efficiency of the power metering device.

[0105] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A smart energy metering method with multiple safety protections, characterized in that, The method includes: S1. Based on the physical structure of the State Grid's new standard metering box, deploy a dual-channel anti-theft sensor network to generate anti-theft sensor layout data; based on the anti-theft sensor layout data, integrate current anomaly detection sensors and magnetic field disturbance monitoring sensors to construct a dual-channel electricity theft behavior perception network. S2. Based on the dual-channel electricity theft behavior perception network, configure multi-parameter edge monitoring nodes, collect multi-source data, and generate multi-source electrical parameter monitoring data; perform edge-end preprocessing on the multi-source electrical parameter monitoring data to generate structured edge monitoring data; S3. Based on structured edge monitoring data, load a lightweight AI diagnostic model to identify electricity theft behavior patterns and diagnose electrical hazards in real time, generating electricity theft risk warning data and equipment health status assessment data; classify the electricity theft risk warning data and equipment health status assessment data into risk levels and generate a four-level warning instruction set; S4. Encrypt and store the four-level early warning instruction set through blockchain nodes to generate an immutable early warning event chain; simultaneously upload structured edge monitoring data to the cloud management platform to build a collaborative data flow between the end, edge, and cloud. S5. Based on the analysis results of the early warning event chain and the cloud management platform, execute proactive operation and maintenance strategies and generate dynamic operation and maintenance task work orders; S6. Based on the operation and maintenance effect data fed back by dynamic operation and maintenance task work orders, iteratively optimize the parameters of the lightweight AI diagnostic model and the blockchain evidence storage strategy to form an adaptive security protection closed-loop system, continuously enhancing the security defense capability and operation and maintenance efficiency of the power metering device.

2. The intelligent management power metering method with multiple safety protections according to claim 1, characterized in that, S1 includes: S11. Analyze the spatial layout of the SMC / PC+ABS non-metallic enclosure to clarify the installation location and distribution pattern of the core components; combine the differences in the number of meters in the metering box to delineate the key coverage area for sensor monitoring and generate anti-theft sensor layout data. S12. Select current anomaly detection sensors and magnetic field disturbance monitoring sensors that are compatible with the rated current range of different metering boxes; plan the installation spacing and fixing method of sensors, disconnect switches and miniature circuit breakers to ensure that the original electrical connection and heat dissipation of the metering box are not affected. S13. Perform parameter calibration on the two sensors to match the detection accuracy with the operating parameter range of the metering box and eliminate the error caused by electromagnetic interference from the non-metallic box. S14. Integrate and install the calibrated sensors according to the layout data to form a dual-channel electricity theft detection network covering the key areas of the enclosure.

3. The intelligent management power metering method with multiple safety protections according to claim 1, characterized in that, The S2 includes: S21. For key components inside the metering box, deploy multi-parameter edge monitoring nodes to synchronously collect current, voltage, temperature, partial discharge and arc characteristic signals through edge monitoring nodes, and summarize them to form multi-source electrical parameter monitoring data; S22. Use a digital filtering algorithm to process the collected data and remove invalid information caused by enclosure vibration and environmental electromagnetic interference. S23. Extract the core features related to electricity theft and equipment failure from the data; S24. Align the time-series data of different monitoring nodes with the time axis, and unify the data acquisition frequency and format standards; integrate the filtered data and core feature quantities to generate structured edge monitoring data.

4. The intelligent management power metering method with multiple safety protections according to claim 1, characterized in that, The S3 includes: S31. Load a lightweight AI diagnostic model adapted to the metering box's operating scenario, and import historical operating data and fault cases of different models of metering boxes; S32. The model uses deep learning to understand the changes in electrical parameters corresponding to different electricity theft behaviors; it also learns the characteristic manifestations of common hidden dangers to improve diagnostic logic. S33. Input the structured edge monitoring data into the model to generate electricity theft risk warning data and equipment health status assessment data; set a four-level risk assessment threshold based on the State Grid safety standards and the metering box rated parameters. S34. Based on the threshold, the early warning data and the evaluation data are classified and judged to generate a four-level early warning instruction set, which includes the handling priority.

5. The intelligent management power metering method with multiple safety protections according to claim 1, characterized in that, The S4 includes: S41. Distribute the four-level early warning instruction set to the distributed blockchain nodes in timestamp order; the blockchain nodes use an asymmetric encryption algorithm to encrypt the instruction set and generate an unalterable encrypted data packet. S42. Connect the encrypted data of each node to form a time-continuous and traceable early warning event chain, and record the early warning time, level and related monitoring data; S43. Extract the core information directly related to measurement accuracy and equipment status from the structured edge monitoring data, and remove redundant data; S44. According to the preset transmission protocol and data format, the core information is synchronously uploaded to the cloud management platform to ensure the stability of data transmission; S45. Establish a real-time communication link between edge monitoring nodes and the cloud management platform, integrate the data flow path of end-side acquisition, edge-side processing, and cloud analysis, and construct a collaborative data flow between the end, edge, and cloud.

6. The intelligent management power metering method with multiple safety protections according to claim 5, characterized in that, S41 includes: S411. Organize the four-level early warning instruction set, extract the generation time information of each instruction, and sort them in chronological order according to timestamps; assign a unique identifier code to each sorted instruction. S412. Analyze the current load status of each node in the distributed blockchain and distribute the sorted instruction set according to the load balancing principle. S413. Each blockchain node receives the corresponding instruction, verifies the instruction identifier code and content integrity, and excludes damaged data. S414. The node initiates an asymmetric encryption algorithm to encrypt the verified instructions and convert them into ciphertext. It then summarizes the encrypted individual instructions from each node and integrates them to form an encrypted data packet.

7. The intelligent management power metering method with multiple safety protections according to claim 6, characterized in that, S412 includes: Collect real-time operational data from each node of the distributed blockchain to generate a node load status dataset; Based on the load status dataset, calculate the current load pressure value of each node and set the load balancing baseline threshold. By comparing the load pressure values ​​of each node with the baseline threshold, the nodes are classified into three levels: light load, medium load, and heavy load. The data volume and processing complexity of a single instruction in the sorted instruction set are statistically analyzed, and the load requirements of each instruction are marked. Match instruction load requirements with node levels, and formulate distribution rules to allocate heavy-load instructions to light-load nodes and medium-load instructions to medium-load nodes; Plan data transmission paths between nodes, avoid overloaded node transmission channels, and accurately distribute the sorted instruction set to the corresponding blockchain nodes according to the rules.

8. The intelligent management power metering method with multiple safety protections according to claim 1, characterized in that, The S5 includes: S51, the cloud management platform receives early warning event chains and structured edge monitoring data, and performs comprehensive analysis by combining the metering box's historical operation and maintenance records with model characteristics; S52. For serious or higher levels of electricity theft risk, automatically trigger a remote control signal to control the opening and closing of the isolating switch and cut off the illegal power circuit. S53. For potential equipment hazards, generate early warning prompts, predict the remaining lifespan of the equipment based on the current transformer operating data, switch operation frequency and temperature change trend, and mark the parts that need to be replaced in advance. S54. Based on the risk level, equipment importance, and maintenance resources, clarify the execution priority and operation process of each maintenance task; S55 integrates remote control commands, potential hazard warnings, lifespan prediction results, and maintenance priorities to generate dynamic maintenance task work orders.

9. The intelligent management power metering method with multiple safety protections according to claim 1, characterized in that, The S6 includes: S61. Collect the execution effect data of dynamic operation and maintenance task work orders, compare the monitoring data before and after operation and maintenance, and analyze the deviation rate, missed judgment rate and false judgment rate of the lightweight AI diagnostic model in electricity theft identification and hidden danger judgment. S62. Based on the analysis results, adjust the internal algorithm parameters of the model, optimize the feature recognition weights of electricity theft and equipment hazards, and optimize the encryption algorithm and data synchronization frequency of blockchain nodes. S63. Based on the actual operation and maintenance results, adjust the risk assessment threshold and early warning classification standards to make the early warning instructions more in line with the on-site operation and maintenance needs; integrate the model parameter optimization results, blockchain evidence storage strategy adjustment scheme and threshold revision standards to form an adaptive security protection closed-loop system.

10. A metering box for implementing the intelligent management power metering method with multiple safety protections as described in claim 1, characterized in that, The metering box includes: Network Deployment Module: Based on the physical structure of the State Grid's new standard metering box, a dual-channel anti-theft sensor network is deployed to generate anti-theft sensor layout data; based on the anti-theft sensor layout data, current anomaly detection sensors and magnetic field disturbance monitoring sensors are integrated to construct a dual-channel electricity theft behavior perception network; Data acquisition module: Based on the dual-channel electricity theft behavior perception network, configure multi-parameter edge monitoring nodes, collect multi-source data, and generate multi-source electrical parameter monitoring data; perform edge-end preprocessing on the multi-source electrical parameter monitoring data to generate structured edge monitoring data; Level Classification Module: Based on structured edge monitoring data, a lightweight AI diagnostic model is loaded to identify electricity theft behavior patterns and diagnose electrical hazards in real time, generating electricity theft risk warning data and equipment health status assessment data; the electricity theft risk warning data and equipment health status assessment data are classified into risk levels, and a four-level warning instruction set is generated; Platform management module: It encrypts and stores the four-level early warning instruction set through blockchain nodes to generate an immutable early warning event chain; it also uploads structured edge monitoring data to the cloud management platform to build a collaborative data flow between the end, edge and cloud. Dynamic Operation and Maintenance Module: Based on the analysis results of the early warning event chain and the cloud management platform, it executes proactive operation and maintenance strategies and generates dynamic operation and maintenance task work orders; Security Defense Module: Based on the operation and maintenance effect data fed back by dynamic operation and maintenance task work orders, the module iteratively optimizes the parameters of the lightweight AI diagnostic model and the blockchain evidence storage strategy to form an adaptive security protection closed-loop system, continuously enhancing the security defense capabilities and operation and maintenance efficiency of the power metering device.

Images