Power grid safety patrol management method and system based on responsibility logic chain tracing and double-loop cooperation

Abstract

The application provides a power grid safety inspection management method and system based on responsibility logic chain tracing and double-loop cooperation, and relates to the technical field of power grid operation and maintenance. According to the equipment fault report, the method determines several defect links of the fault equipment in the safety inspection process and generates rectification tasks for each defect link. For each rectification task, a task process list composed of several task nodes is created. According to the task process list corresponding to each defect link, rectification tasks are issued to each execution personnel, and the task completion status of the task node is monitored. When the unchanged duration of the task completion status of any task node exceeds the preset duration, an alarm information is sent to the task node until the task process list is completed. Therefore, the application can solve the problem of low rectification task completion efficiency due to the lack of real-time supervision of rectification tasks in the prior art.

Description

Technical Field

[0001] This invention relates to the field of power grid operation and maintenance technology, and in particular to a power grid safety inspection and management method and system based on responsibility logic chain tracing and dual-ring collaboration. Background Technology

[0002] In power grid operation and maintenance safety management, apart from sudden environmental factors, equipment failures can be avoided through standardized operation and maintenance procedures, timely inspections and replacement of parts. However, equipment failures may occur due to unreasonable operation and maintenance procedures or poor execution. Therefore, it is common practice to conduct equipment failure analysis to identify defects in each link of the process (such as inspection links, maintenance links, etc.), and then rectify and optimize the operation and maintenance process to reduce the equipment failure rate.

[0003] However, existing technologies have significant shortcomings in rectification. Current mainstream rectification task management relies heavily on manual methods: managers create rectification tasks based on defects in the operation and maintenance process and distribute them via email, telephone, or paper work orders. At the same time, when there are too many rectification tasks, it is difficult to monitor them in real time. After the tasks are distributed, their execution status depends on the voluntary feedback of the executors, which easily leads to low efficiency in completing rectification tasks. Summary of the Invention

[0004] This invention provides a power grid safety inspection and management method and system based on responsibility logic chain tracing and dual-ring collaboration. The method can solve the problem of low efficiency in completing rectification tasks due to the lack of real-time supervision of rectification tasks in the existing technology.

[0005] One embodiment of the present invention provides a power grid safety inspection and management method based on responsibility logic chain tracing and dual-loop collaboration, comprising: Obtain equipment failure reports from faulty devices; Based on the equipment failure report, identify several defective links in the faulty equipment during the safety inspection process and generate rectification tasks for each defective link. For each rectification task, the executors, supervisors, and acceptance personnel are identified, and a task flow list consisting of several task nodes is created based on the executors, supervisors, and acceptance personnel; wherein, the task nodes include: execution nodes, supervision nodes, and acceptance nodes; Based on the task process list corresponding to each defective link, rectification tasks are issued to each of the executors, and the task completion status of each task node is monitored. When the task completion status of any task node is determined to be completed, the next task node is determined and a reminder is issued to the target personnel of the next task node; when the unchanged duration of the task completion status of any task node exceeds a preset time limit, an alarm message is issued to the target personnel of the task node until the task process list is completed; wherein, the target personnel include: executors, supervisors and acceptors.

[0006] Furthermore, the step of determining several defective links in the faulty equipment during the safety inspection process based on the equipment fault report and generating rectification tasks for each defective link includes: Obtain the operation and maintenance process plan for the faulty equipment; Based on the equipment failure report, a defect analysis is performed on each maintenance step in the maintenance process plan to identify several defective steps in the execution of the maintenance process plan and a description of the defects in each defective step. Based on each defective link and its description, a rectification task for each defective link is matched in a pre-set rectification task database. The rectification task database contains pre-set rectification tasks for several defects in each maintenance link.

[0007] Furthermore, based on the equipment failure report, defect analysis is performed on each maintenance step in the maintenance process plan to identify several defective steps in the execution of the maintenance process plan and a description of the defects in each defective step, including: Obtain the inspection work orders and maintenance work orders for the area where the faulty equipment is located; Perform fault analysis on the equipment fault reports to determine the fault characteristics of the faulty equipment; Based on the fault characteristics, inspection rules and maintenance rules are extracted from the operation and maintenance process plan. Based on the inspection rules and maintenance rules, execution defect analysis is performed on the inspection work order and the maintenance work order to determine several defective links and defect descriptions of each defective link.

[0008] Furthermore, the fault characteristics include: a number of faulty parts; the inspection rules include: the planned inspection cycle of the number of faulty parts; the maintenance rules include: the planned maintenance time of the number of faulty parts. The process involves extracting inspection rules and maintenance rules from the operation and maintenance plan, performing execution defect analysis on the inspection work orders and maintenance work orders based on the inspection rules and maintenance rules, identifying several defective links and defect descriptions for each defective link, including: Based on the keywords extracted from the faulty parts in the operation and maintenance process plan, the planned inspection cycle and planned maintenance time of the faulty parts are determined. Based on the faulty part, several actual inspection cycles and actual inspection contents in the inspection work order, select the target inspection cycle related to the faulty part. The planned inspection cycle is matched with the target inspection cycle. When there is a planned inspection cycle that does not match the target inspection cycle, inspection defect information is generated based on the unmatched planned inspection cycle to characterize the failure of the faulty part to be inspected on time. Based on the preset time interval and the planned maintenance time, a record retrieval interval is generated for each faulty part. Based on the record retrieval interval, the maintenance record for each faulty part is retrieved in the maintenance work order. When no maintenance record for any part is found in the record retrieval interval, maintenance defect information is generated to characterize the failure to maintain the faulty part on time, based on the planned maintenance time when no maintenance record is found.

[0009] Furthermore, after identifying several defective links and their defect descriptions, the process also includes: Obtain a preset responsibility tracing rule base; wherein, the responsibility tracing rule base stores several tracing association rules, and the tracing association rules are constructed by using an association rule mining algorithm to analyze the causal relationship strength between historical faulty equipment and each historical defective link based on historical fault analysis data; Based on the fault characteristics and the defective links, semantic similarity matching is performed in the responsibility tracing rule base to determine the target tracing association rules corresponding to each defective link; Based on the target tracing association rules, determine the causal relationship strength between each defective link and the faulty equipment. Based on the strength of the causal relationship, the responsibility ratio of each defective link is calculated, and the defective link with the highest responsibility ratio is identified as the primary defective link.

[0010] An embodiment of the present invention also provides a power grid safety inspection and management system based on responsibility logic chain tracing and dual-loop collaboration, comprising: The report acquisition module is used to acquire equipment failure reports from faulty devices. The task generation module is used to determine several defective links of the faulty equipment during the safety inspection process based on the equipment fault report and generate rectification tasks for each defective link. The list creation module is used to determine the executors, supervisors, and acceptance personnel for each rectification task, and to create a task flow list consisting of several task nodes based on the executors, supervisors, and acceptance personnel respectively; wherein, the task nodes include: execution nodes, supervision nodes, and acceptance nodes. The task monitoring module is used to issue rectification tasks to each of the executors according to the task process list corresponding to each defective link, and to monitor the task completion status of each task node. The task alarm module is used to determine the next task node when the task completion status of any task node is determined to be completed, and to issue a reminder to the target personnel of the next task node; when the unchanged time of the task completion status of any task node exceeds a preset time limit, an alarm message is issued to the target personnel of the task node until the task process list is completed; wherein, the target personnel include: executors, supervisors and acceptance personnel.

[0011] Furthermore, the task generation module, based on the equipment fault report, identifies several defective links in the faulty equipment during the safety inspection process and generates rectification tasks for each defective link, including: Obtain the operation and maintenance process plan for the faulty equipment; Based on the equipment failure report, a defect analysis is performed on each maintenance step in the maintenance process plan to identify several defective steps in the execution of the maintenance process plan and a description of the defects in each defective step. Based on each defective link and its description, a rectification task for each defective link is matched in a pre-set rectification task database. The rectification task database contains pre-set rectification tasks for several defects in each maintenance link.

[0012] Furthermore, the task generation module, based on the equipment fault report, performs defect analysis on each maintenance step in the maintenance process plan to identify several defective steps in the execution of the maintenance process plan and a description of the defects in each defective step, including: Obtain the inspection work orders and maintenance work orders for the area where the faulty equipment is located; Perform fault analysis on the equipment fault reports to determine the fault characteristics of the faulty equipment; Based on the fault characteristics, inspection rules and maintenance rules are extracted from the operation and maintenance process plan. Based on the inspection rules and maintenance rules, execution defect analysis is performed on the inspection work order and the maintenance work order to determine several defective links and defect descriptions of each defective link.

[0013] Furthermore, the fault characteristics include: a number of faulty parts; the inspection rules include: the planned inspection cycle of the number of faulty parts; the maintenance rules include: the planned maintenance time of the number of faulty parts. The task generation module extracts inspection rules and maintenance rules from the operation and maintenance process plan, and performs execution defect analysis on the inspection work order and the maintenance work order according to the inspection rules and maintenance rules to determine several defective links and defect descriptions for each defective link, including: Based on the keywords extracted from the faulty parts in the operation and maintenance process plan, the planned inspection cycle and planned maintenance time of the faulty parts are determined. Based on the faulty part, several actual inspection cycles and actual inspection contents in the inspection work order, select the target inspection cycle related to the faulty part. The planned inspection cycle is matched with the target inspection cycle. When there is a planned inspection cycle that does not match the target inspection cycle, inspection defect information is generated based on the unmatched planned inspection cycle to characterize the failure of the faulty part to be inspected on time. Based on the preset time interval and the planned maintenance time, a record retrieval interval is generated for each faulty part. Based on the record retrieval interval, the maintenance record for each faulty part is retrieved in the maintenance work order. When no maintenance record for any part is found in the record retrieval interval, maintenance defect information is generated to characterize the failure to maintain the faulty part on time, based on the planned maintenance time when no maintenance record is found.

[0014] Furthermore, the power grid safety inspection and management system based on responsibility logic chain tracing and dual-ring collaboration described in the above embodiments also includes: a responsibility tracing module; The responsibility tracing module is used to obtain a preset responsibility tracing rule base after determining several defective links and the defect descriptions of each defective link; wherein, the responsibility tracing rule base stores several tracing association rules, and the tracing association rules are constructed by using an association rule mining algorithm to analyze the causal relationship strength between historical faulty equipment and each historical defective link based on historical fault analysis data; Based on the fault characteristics and the defective links, semantic similarity matching is performed in the responsibility tracing rule base to determine the target tracing association rules corresponding to each defective link; Based on the target tracing association rules, determine the causal relationship strength between each defective link and the faulty equipment. Based on the strength of the causal relationship, the responsibility ratio of each defective link is calculated, and the defective link with the highest responsibility ratio is identified as the primary defective link.

[0015] The following benefits can be obtained by implementing the present invention: This invention provides a power grid safety inspection and management method and system based on responsibility logic chain tracing and dual-loop collaboration. The method, based on equipment fault reports, identifies several defective links in the safety inspection process of the faulty equipment and generates rectification tasks for each defective link. For each rectification task, it identifies executors, supervisors, and acceptance personnel, and creates a task flow list consisting of several task nodes based on these personnel. Each task node includes an execution node, a supervision node, and an acceptance node. Based on the task flow list corresponding to each defective link, rectification tasks are issued to each executor, and the task completion status of each task node is monitored. If the unchanged duration of the task completion status of any task node exceeds a preset time limit, an alarm message is issued for that task node until the task flow list is completed. Therefore, this invention achieves real-time monitoring of rectification tasks by creating task lists to remind or alarm the target personnel at task nodes, thereby improving rectification efficiency. Attached Figure Description

[0016] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0017] Figure 1 This is a flowchart illustrating a power grid safety inspection and management method based on responsibility logic chain tracing and dual-ring collaboration, provided in a certain embodiment of this application. Figure 2 This is a schematic diagram of the structure of a power grid safety inspection and management system based on responsibility logic chain tracing and dual-ring collaboration, provided in a certain embodiment of this application. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0020] In the description of the embodiments of this application, technical terms such as "first" and "second" are used only to distinguish different objects and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly defined.

[0021] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0022] In the description of the embodiments in this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0023] In the description of the embodiments of this application, the term "multiple" refers to two or more (including two), similarly, "multiple sets" refers to two or more (including two sets), and "multiple pieces" refers to two or more (including two pieces).

[0024] In the description of the embodiments of this application, unless otherwise expressly specified and limited, technical terms such as "installation," "connection," "joining," and "fixing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. For those skilled in the art, the specific meaning of the above terms in the embodiments of this application can be understood according to the specific circumstances.

[0025] See Figure 1 To address the problems in the prior art, an embodiment of the present invention provides a power grid safety inspection and management method based on responsibility logic chain tracing and dual-ring collaboration, comprising: S1. Obtain the equipment failure report of the faulty equipment; In a preferred embodiment of the present invention, after a device malfunctions, a device malfunction report is generated by relevant personnel or by a malfunction detection model, including: malfunction time, malfunction phenomenon, device information, etc.

[0026] S2. Based on the equipment failure report, identify several defective links of the faulty equipment during the safety inspection process and generate rectification tasks for each defective link. Preferably, the step of determining several defective links in the faulty equipment during the safety inspection process based on the equipment fault report and generating rectification tasks for each defective link includes: Obtain the operation and maintenance process plan for the faulty equipment; Based on the equipment failure report, a defect analysis is performed on each maintenance step in the maintenance process plan to identify several defective steps in the execution of the maintenance process plan and a description of the defects in each defective step. Based on each defective link and its description, a rectification task for each defective link is matched in a pre-set rectification task database. The rectification task database contains pre-set rectification tasks for several defects in each maintenance link.

[0027] Preferably, the step of performing defect analysis on each maintenance step in the maintenance process plan based on the equipment failure report, and identifying several defective steps in the execution of the maintenance process plan and a defect description for each defective step, includes: Obtain the inspection work orders and maintenance work orders for the area where the faulty equipment is located; Perform fault analysis on the equipment fault reports to determine the fault characteristics of the faulty equipment; Based on the fault characteristics, inspection rules and maintenance rules are extracted from the operation and maintenance process plan. Based on the inspection rules and maintenance rules, execution defect analysis is performed on the inspection work order and the maintenance work order to determine several defective links and defect descriptions of each defective link.

[0028] Preferably, the fault characteristics include: a number of faulty parts; the inspection rules include: the planned inspection cycle of the number of faulty parts; and the maintenance rules include: the planned maintenance time of the number of faulty parts. The process involves extracting inspection rules and maintenance rules from the operation and maintenance plan, performing execution defect analysis on the inspection work orders and maintenance work orders based on the inspection rules and maintenance rules, identifying several defective links and defect descriptions for each defective link, including: Based on the keywords extracted from the faulty parts in the operation and maintenance process plan, the planned inspection cycle and planned maintenance time of the faulty parts are determined. Based on the faulty part, several actual inspection cycles and actual inspection contents in the inspection work order, select the target inspection cycle related to the faulty part. The planned inspection cycle is matched with the target inspection cycle. When there is a planned inspection cycle that does not match the target inspection cycle, inspection defect information is generated based on the unmatched planned inspection cycle to characterize the failure of the faulty part to be inspected on time. Based on the preset time interval and the planned maintenance time, a record retrieval interval is generated for each faulty part. Based on the record retrieval interval, the maintenance record for each faulty part is retrieved in the maintenance work order. When no maintenance record for any part is found in the record retrieval interval, maintenance defect information is generated to characterize the failure to maintain the faulty part on time, based on the planned maintenance time when no maintenance record is found.

[0029] In a preferred embodiment of the present invention, three core features are extracted based on the equipment fault report: fault phenomenon, equipment location, and fault type. Taking a transformer oil leakage case as an example, the extracted features are: fault phenomenon = "oil traces at the bottom seal of the oil tank, oil level 30mm lower than the standard value"; equipment location = substation #2 main transformer; fault type = equipment sealing defect.

[0030] Based on the faulty component—the seal—the inspection and maintenance work orders for the equipment were retrieved, with a focus on verifying the operation records related to the seal. The verification included the seal replacement plan and inspection process records. If the service life (planned maintenance time) of the transformer seal was found to be 3 years, and the maintenance records for the seal were retrieved from the maintenance work orders of the third year, and no maintenance records for replacement parts were found within the third year, and the planned inspection cycle was weekly (meaning the seal should be inspected weekly), and if the seal inspection item was blank in the weekly inspection record, then both the maintenance and inspection processes were confirmed to be defective. The defects were described as part maintenance not performed and inspection omissions, respectively.

[0031] Therefore, it is crucial to identify the responsibilities of relevant personnel in the maintenance and inspection processes. For example, if maintenance personnel fail to tighten bolts according to procedures, it constitutes a failure of responsibility at the execution level; if the technical supervisor fails to update maintenance process standards in a timely manner or the team leader fails to properly inspect maintenance quality, it constitutes a failure of responsibility at the management level; and if the safety specialist fails to supervise and conduct spot checks on the completeness of inspection work, it constitutes a failure of responsibility at the supervisory level.

[0032] To address the deficiencies in the operation and maintenance process and the lack of personnel responsibility, relevant personnel were trained and the operation and maintenance process was rectified. Specifically, through the outer-loop rectification collaboration mechanism, the deficient links were linked to the pre-set rectification measures database. Through keyword matching algorithms, the content of the deficient links was intelligently matched with the solutions in the rectification task database to generate a targeted rectification task list.

[0033] The specific rectification task database stores standard rectification plans pre-set by relevant management personnel for various defect scenarios. For example, for defects in maintenance processes not performed according to the operation and maintenance plan, rectification tasks include retraining and assessing relevant personnel. For defects caused by execution errors, rectification tasks include correcting the errors. For instance, in the operation and maintenance process, for example, regarding excessive contact resistance of contacts, standard maintenance procedures include power outage inspection, cleaning contacts, applying conductive grease, and tightening to the standard torque. However, when it is detected that maintenance personnel have not strictly followed the standard procedures, the rectification task for that maintenance process includes organizing procedure training and assessment for maintenance personnel.

[0034] Preferably, after determining several defective links and the defect description of each defective link, the method further includes: Obtain a preset responsibility tracing rule base; wherein, the responsibility tracing rule base stores several tracing association rules, and the tracing association rules are constructed by using an association rule mining algorithm to analyze the causal relationship strength between historical faulty equipment and each historical defective link based on historical fault analysis data; Based on the fault characteristics and the defective links, semantic similarity matching is performed in the responsibility tracing rule base to determine the target tracing association rules corresponding to each defective link; Based on the target tracing association rules, determine the causal relationship strength between each defective link and the faulty equipment. Based on the strength of the causal relationship, the responsibility ratio of each defective link is calculated, and the defective link with the highest responsibility ratio is identified as the primary defective link.

[0035] Furthermore, when defects exist in the operation and maintenance process, this embodiment also assigns responsibility to relevant personnel. Specifically, based on historical fault reports and historical defect analysis reports of various power equipment, the Apriori association rule algorithm (association rule mining algorithm) and causal association verification are used to construct an association rule base, thereby locating the defect. The Apriori algorithm is used to mine high-frequency association itemsets between multi-dimensional data, while causal association verification is used to eliminate false associations, strengthen causal logic, and adapt to the causal chains in the operation and maintenance process that lead to process failures and the resulting lack of responsibility, ultimately forming a reliable association rule base.

[0036] Specifically, the steps for association mining, i.e., constructing association rules, are as follows: using historical fault reports and historical defect analysis reports of power grid equipment from the past three years as the mining data, the data includes equipment fault records, process execution records, and responsibility determination records: First, historical data preprocessing and transaction sets are constructed. The historical power grid data undergoes a structured transformation to create transaction sets adapted to the Apriori algorithm. Historical data is categorized into three main dimensions: equipment defects (A), process failures (B), and lack of accountability (C). Each dimension is further subdivided into specific entities. For example, category A includes 20 types of equipment defects such as aging seals (A1) and insulation damage (A2); category B includes 15 types of process failures such as failure to execute periodic replacement procedures (B1) and omissions in inspection procedures (B2); and category C includes 12 types of lack of accountability such as missed inspections by inspectors (C1) and inadequate supervision by management personnel (C2).

[0037] Each transaction is based on a single fault event as the smallest unit of transaction, and each transaction contains a combination of A, B, and C type entities corresponding to that event. For example, the transaction record for a transformer oil leakage event is {Aging of seals A1, failure to execute regular replacement process B1, missed inspection by inspector C1, inadequate supervision by management personnel C2}; Secondly, by leveraging the Apriori algorithm, high-frequency coexistence combinations across the three dimensions are identified. Based on the data distribution, a minimum support of 5% is set, meaning that a combination is considered a high-frequency combination if its frequency of occurrence in all transactions is ≥5%. The transaction set is traversed, and the frequency of occurrence of individual entities is counted to select first-order frequent itemsets with support ≥5%. For first-order frequent itemsets, second-order and third-order candidate sets are generated layer by layer. Invalid candidate sets are eliminated through a pruning strategy (if a subset of a candidate set is not a frequent itemset, then the candidate set must be not a frequent itemset). Finally, high-order frequent itemsets with support ≥5% are selected.

[0038] Then, based on the mined frequent itemsets, association rules between preceding and succeeding items are generated. Valid rules are then filtered using confidence levels. Causal association rules are extracted from higher-order frequent itemsets, and combined with power grid management logic, the rules are categorized as "Category A → Category B" and "Category B → Category C". For example, rule A1 → B1 generated from frequent itemset {A1, B1} indicates that the seal is aging → the periodic replacement process was not executed; rule B1 → C1 generated from frequent itemset {B1, C1} indicates that the periodic replacement process was not executed → the inspector missed an inspection. Confidence level = support (preceding item ∪ succeeding item) / support (preceding item), used to measure the probability of the succeeding item occurring when the preceding item occurs. Considering the accuracy requirements for accountability, a minimum confidence level of 70% is set, and valid rules with a confidence level ≥ 70% are selected.

[0039] Furthermore, a Bayesian network causal inference method is used to eliminate spurious associations through causal verification: a network is constructed containing entities of classes A, B, and C, as well as potential interfering factors, such as environmental humidity. Historical data is input to train network parameters. The average causal effect (ACE) is calculated to quantify the causal influence of the preceding term on the following term. The causal strength threshold is set to 0.6 (the value ranges from 0 to 1, and the closer it is to 1, the stronger the causal relationship). If a rule has a confidence level that meets the threshold but a causal strength < 0.6, it is judged as a spurious association. For example, if the equipment has been in operation for too long, it will cause both A2 and B3. Since there is no direct causal relationship between the two, the rule will be removed. Rules with a causal strength ≥ 0.6 are retained as the final valid rules.

[0040] Finally, the filtered valid association rules are stored according to the categories of equipment defects → process failure and process failure → lack of responsibility, and an association rule library is constructed. The specific content includes: rule ID, preceding item type, preceding item name, following item type, following item name, confidence level, causal strength, and applicable equipment type. At the same time, the rule library supports dynamic updates to optimize the confidence level and causal strength of the rules and improve the adaptability of the logical chain analysis model.

[0041] In practical applications, by calling the mapping rules from abnormal features to equipment defects in the association rule base, the extracted core features are semantically matched with the feature descriptions of equipment defects (Class A) in the rule base. The similarity is calculated, and the threshold is set to 75%. Taking the seal failure as an example, the features of oil stains at the seal, low oil level, and seal-related defects have a similarity of 82% with the feature description of A1 (seal aging) in the rule base, and a preliminary match is made to the candidate equipment defect node A1 - seal aging.

[0042] By synchronously retrieving on-site inspection data from the equipment, including high-definition images of the seal taken by the intelligent inspection instrument and disassembly pre-inspection records, the authenticity of candidate defects is verified. If the image features show cracks and decreased elasticity on the surface of the sealing gasket, and the disassembly pre-inspection records show that the gasket hardness exceeds the standard range, the candidate defect node is confirmed to be valid. If the inspection data does not match the candidate defect, such as if the sealing gasket is intact, a secondary matching is triggered to re-screen other equipment defect rules. After feature matching and data verification, the equipment defect node is finally determined to be the aging and elasticity failure of the sealing gasket at the bottom of the oil tank (A1), which is the direct physical cause of the abnormal problem.

[0043] Furthermore, valid rules for equipment defects → process failures are retrieved from the association rule base. The identified equipment defect nodes are used as the preceding terms to match the corresponding subsequent process failure types. For example, rule R001 in the rule base (A1→B1, confidence level 84.7%, causality strength 0.72) clearly shows a strong causal relationship between aging seals (A1) and failure of the periodic replacement process (B1). The model initially matches the candidate process failure node B1 - failure of the periodic replacement process.

[0044] By retrieving the entire lifecycle execution records of the equipment, the focus is on verifying the operation records related to the candidate process failure nodes. Taking B1 as an example, the verification content includes: seal replacement plan, plan execution records, and inspection process records. If the verification finds that the transformer seal has been used continuously for 5 years without any replacement operation documents, and the seal inspection items in the most recent inspection records are all blank, then the candidate process failure node is verified as valid, confirming the existence of process loopholes such as the failure to execute the regular replacement process and the omission of items in the inspection process.

[0045] If there are multiple associated process failure types, then based on the causal strength sorting, all process failure types with a causal strength ≥ 0.6 are included in the logic tree to form structured process failure sub-nodes.

[0046] Furthermore, the system retrieves valid rules from the association rule base indicating process failure → lack of responsibility, and uses each process failure node as a preceding term to match the corresponding subsequent type of lack of responsibility. For example, process failure node B1 matches rule C2 with a confidence level of 79.2% and a causal strength of 0.68, corresponding to inadequate supervision by management personnel (C2).

[0047] By retrieving the job descriptions for power grid operation and maintenance positions, the core responsibilities of the execution, management, and supervisory levels were clarified, and the responsible parties were further verified by combining this information with personnel operation records. Regarding the inspector's failure to inspect C1: The monthly inspection responsibility allocation record for the equipment was checked, confirming that inspector Zhang was the designated inspector for the equipment recently, and that the seal inspection items in his inspection record were all blank, with no on-site sign-in confirmation record, verifying that Zhang was the directly responsible party at the execution level; Regarding the inadequate supervision by management personnel (C2): An investigation of the process supervision records of Li, the team leader of the maintenance team, revealed that he failed to review the service life of the sealing components every quarter as required by the system, and also failed to urge the inspectors to rectify the blank inspection items. This verifies that Li is an indirect responsible party of the management. Based on the power grid safety supervision responsibility list, the Safety Supervision Department is required to conduct a special supervision of the operation and maintenance process every six months. The review found that the substation's seal replacement process had not been supervised recently. Matching rule B1→C3 was added, with a confidence level of 72.1% and a causal strength of 0.63, confirming a lack of supervision and inspection at the supervisory level, specifically C3. Personnel responsibility factors were determined: based on the rule matching and record verification results, the final responsibility deficiencies were identified as C1 - Inspector Zhang's missed inspection, C2 - Inadequate supervision by the operation and maintenance team leader Li, and C3 - Lack of supervision and inspection by the Safety Supervision Department. The responsible parties, types of responsibility, and legal basis for each node were clearly defined.

[0048] Finally, an inner-loop responsibility tracing mechanism is adopted. Based on the constructed logic tree, especially the sub-nodes of missing responsibility and process failure, reverse tracing is performed along the predefined three-level responsibility chain of execution layer → management layer → supervision layer. Using the preset responsibility boundaries of each level as a benchmark, the process failure sub-nodes are matched to the responsibility scope of the corresponding level, and then the specific responsible entity at that level is identified through the sub-nodes of missing responsibility, ultimately forming a complete location link of process failure sub-node → level responsibility matching → missing responsibility sub-node → specific responsible person.

[0049] Execution layer: The associated process failure sub-nodes are omissions in the inspection process (B2) and violations of basic operation procedures; the associated responsibility deficiency sub-nodes are omissions by inspectors (C1) and non-standard operations. Management: The related process failure sub-nodes are failure to implement the regularly changed process (B1), lack of planning and inadequate training process; the related responsibility failure sub-nodes are inadequate supervision by management personnel (C2) and failure to track the plan.

[0050] Supervisory layer: The associated process failure sub-nodes are all process failure sub-nodes, and the associated responsibility deficiency sub-nodes are the lack of supervision and inspection by the supervisory layer (C3) and the failure to hold people accountable in a timely manner.

[0051] The connection between each level and its sub-nodes is essentially a causal chain of responsibility performance status → process execution result → responsibility determination. The lack of responsibility at the execution and management levels directly leads to the generation of corresponding process failure sub-nodes. The lack of responsibility at the supervisory level allows process failure sub-nodes to continue to exist. The responsibility-deficient sub-nodes are a precise characterization of the lack of responsibility at each level, ultimately achieving the goal of tracing back to the matter, management, people, and responsibilities.

[0052] Finally, the fuzzy comprehensive evaluation method was used to determine the responsibility weights at each level. The specific steps are as follows: Determine the evaluation factor set: U = {Execution level responsibility U1, Management level responsibility U2, Supervisory level responsibility U3}; Determine the weight set: Based on expert scores (3 power grid safety management experts), the initial weights were calculated using the analytic hierarchy process (AHP), and after consistency verification, the weight set W = {0.4, 0.35, 0.25} was determined; Determine the comment set: V = {Main responsibility, Secondary responsibility, General responsibility}, with corresponding quantitative values ​​of {0.8, 0.5, 0.2}; Calculate the comprehensive evaluation result: Through fuzzy matrix operations, the evaluation result for execution level responsibility was 0.75 (main responsibility), for management level responsibility it was 0.6 (secondary responsibility), and for supervisory level responsibility it was 0.3 (general responsibility); The comprehensive evaluation result was calculated by combining the evaluation factor set U, the weight set W, and the comment set V using fuzzy matrix operations, as follows: For each element in the evaluation factor set U, determine its membership degree to the elements in each comment set. Take the average expert scores to form an n×m matrix R, where n is the number of elements in U and m is the number of elements in V, reflecting the degree of correlation between the two. Calculate B=[b1,b2,…,bm] using the formula B=W・R (matrix multiplication), representing the comprehensive membership degree of all elements in the evaluation factor set U to the elements in the comment set.

[0053] Since the elements in the evaluation factor set U in the power grid responsibility tracing have clear responsibility directions, the single-factor evaluation vector × V quantization value is used first (the element Ui in the evaluation factor set U corresponds to the row vector V quantization value of R); it can also be calculated through B·V quantization value. The final result matches the threshold of the comment set, and when the value is ≥ threshold 0.7, it is set as the main responsibility, and the evaluation is completed.

[0054] S3. For each rectification task, determine the executors, supervisors, and acceptance personnel, and create a task flow list consisting of several task nodes based on the executors, supervisors, and acceptance personnel; wherein, the task nodes include: execution nodes, supervision nodes, and acceptance nodes. S4. Based on the task flow list corresponding to each defective link, issue rectification tasks to each executor and monitor the task completion status of each task node; S5. When the task completion status of any task node is determined to be completed, the next task node is determined and a reminder is issued to the target personnel of the next task node; when the unchanged duration of the task completion status of any task node exceeds a preset time limit, an alarm message is issued to the target personnel of the task node until the task process list is completed; wherein, the target personnel include: executors, supervisors and acceptance personnel.

[0055] In a preferred embodiment of the present invention, the inspection and maintenance personnel of the faulty equipment are designated as the corresponding executors. The superior managers of these executors are determined from a job list and designated as supervisors. The safety management personnel of the area where the faulty equipment is located are designated as acceptance personnel. Taking the rectification task as an example, and retraining and assessing relevant personnel, after receiving the rectification task, the executors complete the task by uploading their training and assessment records. The task is then transferred to the next supervisory node for preliminary review, and finally, the acceptance node reviews the submitted supporting materials. If the acceptance is successful, the issue status is marked as closed, and all relevant materials are archived. If the acceptance fails, the reviewer fills in rejection comments, and the task is automatically reassigned to the executor node, requiring optimization and rectification until final acceptance. Simultaneously, the task nodes are monitored, and overdue alarms are automatically triggered for tasks that are about to expire or have already expired.

[0056] Specifically, after receiving the task, the personnel responsible for implementation upload rectification process data in real time via a mobile APP: Execution level task (T2024061501): Zhang uploads on-site photos of the sealing gasket replacement (3 photos, including removal of the old part, installation of the new part, and oil level replenishment), and a pressure test report (showing a test pressure of 0.03MPa, stable for 30 minutes with no leakage); Management level task (T2024061502): Li uploads the revised process documents, training attendance sheet, and assessment results sheet (all staff passed); By automatically tracking the task progress, for tasks that are not submitted with 20% remaining before the deadline (e.g., 12 hours remaining for T2024061501), an overdue warning is sent to the responsible person and the next level of management. After rectification, verification is conducted in a hierarchical manner: self-inspection at the execution level → review at the management level → acceptance at the supervisory level. Self-inspection at the execution level: After completing the replacement, Zhang checks the sealing status, oil level, and pressure data, and submits a self-inspection report confirming compliance. Review at the management level: Team leader Li verifies the rectification on-site, checks the process data against the self-inspection report, and signs the review opinion. Acceptance at the supervisory level: Wang from the safety supervision department conducts random checks on the rectification results of all tasks, verifies the equipment status on-site, reviews process documents and training records, and signs the acceptance opinion. If the acceptance fails, the system returns the task to the responsible person, noting the rectification defects, requiring re-rectification and resubmission within 3 working days. The visual interface integrates an inner loop of responsibility traceability and an outer loop of rectification collaboration: The inner loop displays a three-level responsibility chain diagram, marking each responsible person and their responsibility weight, the logic tree decomposition process, and key information from the responsibility traceability report; the outer loop displays a rectification task progress bar, categorized by completion rate and overdue rate, task review and transfer records, and evidence to verify rectification results; managers can view real-time data via PC, supporting filtering and querying by device number, time range, and responsibility level, achieving full-chain traceability and monitoring.

[0057] See Figure 2 This invention provides a power grid safety inspection and management system based on responsibility logic chain tracing and dual-ring collaboration, comprising: The report acquisition module is used to acquire equipment failure reports from faulty devices. In a preferred embodiment of the present invention, after a device malfunctions, the report acquisition module obtains a device malfunction report generated by relevant personnel or a malfunction detection model, which includes: malfunction time, malfunction phenomenon, device information, etc.

[0058] The task generation module is used to determine several defective links of the faulty equipment during the safety inspection process based on the equipment fault report and generate rectification tasks for each defective link. In a preferred embodiment of the present invention, three core features are extracted based on the equipment fault report: fault phenomenon, equipment location, and fault type. Taking a transformer oil leakage case as an example, the extracted features are: fault phenomenon = "oil traces at the bottom seal of the oil tank, oil level 30mm lower than the standard value"; equipment location = substation #2 main transformer; fault type = equipment sealing defect.

[0059] Based on the faulty component—the seal—the inspection and maintenance work orders for the equipment were retrieved, with a focus on verifying the operation records related to the seal. The verification included the seal replacement plan and inspection process records. If the service life (planned maintenance time) of the transformer seal was found to be 3 years, and the maintenance records for the seal were retrieved from the maintenance work orders of the third year, and no maintenance records for replacement parts were found within the third year, and the planned inspection cycle was weekly (meaning the seal should be inspected weekly), and if the seal inspection item was blank in the weekly inspection record, then both the maintenance and inspection processes were confirmed to be defective. The defects were described as part maintenance not performed and inspection omissions, respectively.

[0060] Therefore, it is crucial to identify the responsibilities of relevant personnel in the maintenance and inspection processes. For example, if maintenance personnel fail to tighten bolts according to procedures, it constitutes a failure of responsibility at the execution level; if the technical supervisor fails to update maintenance process standards in a timely manner or the team leader fails to properly inspect maintenance quality, it constitutes a failure of responsibility at the management level; and if the safety specialist fails to supervise and conduct spot checks on the completeness of inspection work, it constitutes a failure of responsibility at the supervisory level.

[0061] To address the deficiencies in the operation and maintenance process and the lack of personnel responsibility, relevant personnel were trained and the operation and maintenance process was rectified. Specifically, through the outer-loop rectification collaboration mechanism, the deficient links were linked to the pre-set rectification measures database. Through keyword matching algorithms, the content of the deficient links was intelligently matched with the solutions in the rectification task database to generate a targeted rectification task list.

[0062] The specific rectification task database stores standard rectification plans pre-set by relevant management personnel for various defect scenarios. For example, for defects in maintenance processes not performed according to the operation and maintenance plan, rectification tasks include retraining and assessing relevant personnel. For defects caused by execution errors, rectification tasks include correcting the errors. For instance, in the operation and maintenance process, for example, regarding excessive contact resistance of contacts, standard maintenance procedures include power outage inspection, cleaning contacts, applying conductive grease, and tightening to the standard torque. However, when it is detected that maintenance personnel have not strictly followed the standard procedures, the rectification task for that maintenance process includes organizing procedure training and assessment for maintenance personnel.

[0063] The list creation module is used to determine the executors, supervisors, and acceptance personnel for each rectification task, and to create a task flow list consisting of several task nodes based on the executors, supervisors, and acceptance personnel respectively; wherein, the task nodes include: execution nodes, supervision nodes, and acceptance nodes. The task monitoring module is used to issue rectification tasks to each of the executors according to the task process list corresponding to each defective link, and to monitor the task completion status of each task node. The task alarm module is used to determine the next task node when the task completion status of any task node is determined to be completed, and to issue a reminder to the target personnel of the next task node; when the unchanged time of the task completion status of any task node exceeds a preset time limit, an alarm message is issued to the target personnel of the task node until the task process list is completed; wherein, the target personnel include: executors, supervisors and acceptance personnel.

[0064] Preferably, the task generation module, based on the equipment fault report, determines several defective links in the faulty equipment during the safety inspection process and generates rectification tasks for each defective link, including: Obtain the operation and maintenance process plan for the faulty equipment; Based on the equipment failure report, a defect analysis is performed on each maintenance step in the maintenance process plan to identify several defective steps in the execution of the maintenance process plan and a description of the defects in each defective step. Based on each defective link and its description, a rectification task for each defective link is matched in a pre-set rectification task database. The rectification task database contains pre-set rectification tasks for several defects in each maintenance link.

[0065] Preferably, the task generation module, based on the equipment fault report, performs defect analysis on each maintenance step in the maintenance process plan to identify several defective steps in the execution of the maintenance process plan and a description of the defects in each defective step, including: Obtain the inspection work orders and maintenance work orders for the area where the faulty equipment is located; Perform fault analysis on the equipment fault reports to determine the fault characteristics of the faulty equipment; Based on the fault characteristics, inspection rules and maintenance rules are extracted from the operation and maintenance process plan. Based on the inspection rules and maintenance rules, execution defect analysis is performed on the inspection work order and the maintenance work order to determine several defective links and defect descriptions of each defective link.

[0066] Preferably, the fault characteristics include: a number of faulty parts; the inspection rules include: the planned inspection cycle of the number of faulty parts; and the maintenance rules include: the planned maintenance time of the number of faulty parts. The task generation module extracts inspection rules and maintenance rules from the operation and maintenance process plan, and performs execution defect analysis on the inspection work order and the maintenance work order according to the inspection rules and maintenance rules to determine several defective links and defect descriptions for each defective link, including: Based on the keywords extracted from the faulty parts in the operation and maintenance process plan, the planned inspection cycle and planned maintenance time of the faulty parts are determined. Based on the faulty part, several actual inspection cycles and actual inspection contents in the inspection work order, select the target inspection cycle related to the faulty part. The planned inspection cycle is matched with the target inspection cycle. When there is a planned inspection cycle that does not match the target inspection cycle, inspection defect information is generated based on the unmatched planned inspection cycle to characterize the failure of the faulty part to be inspected on time. Based on the preset time interval and the planned maintenance time, a record retrieval interval is generated for each faulty part. Based on the record retrieval interval, the maintenance record for each faulty part is retrieved in the maintenance work order. When no maintenance record for any part is found in the record retrieval interval, maintenance defect information is generated to characterize the failure to maintain the faulty part on time, based on the planned maintenance time when no maintenance record is found.

[0067] Furthermore, the power grid safety inspection and management system based on responsibility logic chain tracing and dual-ring collaboration described in the above embodiments also includes: a responsibility tracing module; The responsibility tracing module is used to obtain a preset responsibility tracing rule base after determining several defective links and the defect descriptions of each defective link; wherein, the responsibility tracing rule base stores several tracing association rules, and the tracing association rules are constructed by using an association rule mining algorithm to analyze the causal relationship strength between historical faulty equipment and each historical defective link based on historical fault analysis data; Based on the fault characteristics and the defective links, semantic similarity matching is performed in the responsibility tracing rule base to determine the target tracing association rules corresponding to each defective link; Based on the target tracing association rules, determine the causal relationship strength between each defective link and the faulty equipment. Based on the strength of the causal relationship, the responsibility ratio of each defective link is calculated, and the defective link with the highest responsibility ratio is identified as the primary defective link.

[0068] Furthermore, when defects exist in the operation and maintenance process, this embodiment also uses a responsibility tracing module to assign responsibility to relevant personnel. Specifically, based on historical fault reports and historical defect analysis reports of various power equipment, the Apriori association rule algorithm (association rule mining algorithm) and causal association verification are used to construct an association rule base, thereby locating the defect. The Apriori algorithm is used to mine high-frequency association itemsets between multi-dimensional data, while causal association verification is used to eliminate false associations, strengthen causal logic, and adapt to the causal chains in the operation and maintenance process that lead to process failure and subsequent lack of responsibility, ultimately forming a reliable association rule base.

[0069] Specifically, the steps for association mining, i.e., constructing association rules, are as follows: using historical fault reports and historical defect analysis reports of power grid equipment from the past three years as the mining data, the data includes equipment fault records, process execution records, and responsibility determination records: First, historical data preprocessing and transaction sets are constructed. The historical power grid data undergoes a structured transformation to create transaction sets adapted to the Apriori algorithm. Historical data is categorized into three main dimensions: equipment defects (A), process failures (B), and lack of accountability (C). Each dimension is further subdivided into specific entities. For example, category A includes 20 types of equipment defects such as aging seals (A1) and insulation damage (A2); category B includes 15 types of process failures such as failure to execute periodic replacement procedures (B1) and omissions in inspection procedures (B2); and category C includes 12 types of lack of accountability such as missed inspections by inspectors (C1) and inadequate supervision by management personnel (C2).

[0070] Each transaction is based on a single fault event as the smallest unit of transaction, and each transaction contains a combination of A, B, and C type entities corresponding to that event. For example, the transaction record for a transformer oil leakage event is {Aging of seals A1, failure to execute regular replacement process B1, missed inspection by inspector C1, inadequate supervision by management personnel C2}; Secondly, by leveraging the Apriori algorithm, high-frequency coexistence combinations across the three dimensions are identified. Based on the data distribution, a minimum support of 5% is set, meaning that a combination is considered a high-frequency combination if its frequency of occurrence in all transactions is ≥5%. The transaction set is traversed, and the frequency of occurrence of individual entities is counted to select first-order frequent itemsets with support ≥5%. For first-order frequent itemsets, second-order and third-order candidate sets are generated layer by layer. Invalid candidate sets are eliminated through a pruning strategy (if a subset of a candidate set is not a frequent itemset, then the candidate set must be not a frequent itemset). Finally, high-order frequent itemsets with support ≥5% are selected.

[0071] Then, based on the mined frequent itemsets, association rules between preceding and succeeding items are generated. Valid rules are then filtered using confidence levels. Causal association rules are extracted from higher-order frequent itemsets, and combined with power grid management logic, the rules are categorized as "Category A → Category B" and "Category B → Category C". For example, rule A1 → B1 generated from frequent itemset {A1, B1} indicates that the seal is aging → the periodic replacement process was not executed; rule B1 → C1 generated from frequent itemset {B1, C1} indicates that the periodic replacement process was not executed → the inspector missed an inspection. Confidence level = support (preceding item ∪ succeeding item) / support (preceding item), used to measure the probability of the succeeding item occurring when the preceding item occurs. Considering the accuracy requirements for accountability, a minimum confidence level of 70% is set, and valid rules with a confidence level ≥ 70% are selected.

[0072] Furthermore, a Bayesian network causal inference method is used to eliminate spurious associations through causal verification: a network is constructed containing entities of classes A, B, and C, as well as potential interfering factors, such as environmental humidity. Historical data is input to train network parameters. The average causal effect (ACE) is calculated to quantify the causal influence of the preceding term on the following term. The causal strength threshold is set to 0.6 (the value ranges from 0 to 1, and the closer it is to 1, the stronger the causal relationship). If a rule has a confidence level that meets the threshold but a causal strength < 0.6, it is judged as a spurious association. For example, if the equipment has been in operation for too long, it will cause both A2 and B3. Since there is no direct causal relationship between the two, the rule will be removed. Rules with a causal strength ≥ 0.6 are retained as the final valid rules.

[0073] Finally, the filtered valid association rules are stored according to the categories of equipment defects → process failure and process failure → lack of responsibility, and an association rule library is constructed. The specific content includes: rule ID, preceding item type, preceding item name, following item type, following item name, confidence level, causal strength, and applicable equipment type. At the same time, the rule library supports dynamic updates to optimize the confidence level and causal strength of the rules and improve the adaptability of the logical chain analysis model.

[0074] In practical applications, by calling the mapping rules from abnormal features to equipment defects in the association rule base, the extracted core features are semantically matched with the feature descriptions of equipment defects (Class A) in the rule base. The similarity is calculated, and the threshold is set to 75%. Taking the seal failure as an example, the features of oil stains at the seal, low oil level, and seal-related defects have a similarity of 82% with the feature description of A1 (seal aging) in the rule base, and a preliminary match is made to the candidate equipment defect node A1 - seal aging.

[0075] By synchronously retrieving on-site inspection data from the equipment, including high-definition images of the seal taken by the intelligent inspection instrument and disassembly pre-inspection records, the authenticity of candidate defects is verified. If the image features show cracks and decreased elasticity on the surface of the sealing gasket, and the disassembly pre-inspection records show that the gasket hardness exceeds the standard range, the candidate defect node is confirmed to be valid. If the inspection data does not match the candidate defect, such as if the sealing gasket is intact, a secondary matching is triggered to re-screen other equipment defect rules. After feature matching and data verification, the equipment defect node is finally determined to be the aging and elasticity failure of the sealing gasket at the bottom of the oil tank (A1), which is the direct physical cause of the abnormal problem.

[0076] Furthermore, valid rules for equipment defects → process failures are retrieved from the association rule base. The identified equipment defect nodes are used as the preceding terms to match the corresponding subsequent process failure types. For example, rule R001 in the rule base (A1→B1, confidence level 84.7%, causality strength 0.72) clearly shows a strong causal relationship between aging seals (A1) and failure of the periodic replacement process (B1). The model initially matches the candidate process failure node B1 - failure of the periodic replacement process.

[0077] By retrieving the entire lifecycle execution records of the equipment, the focus is on verifying the operation records related to the candidate process failure nodes. Taking B1 as an example, the verification content includes: seal replacement plan, plan execution records, and inspection process records. If the verification finds that the transformer seal has been used continuously for 5 years without any replacement operation documents, and the seal inspection items in the most recent inspection records are all blank, then the candidate process failure node is verified as valid, confirming the existence of process loopholes such as the failure to execute the regular replacement process and the omission of items in the inspection process.

[0078] If there are multiple associated process failure types, then based on the causal strength sorting, all process failure types with a causal strength ≥ 0.6 are included in the logic tree to form structured process failure sub-nodes.

[0079] Furthermore, the system retrieves valid rules from the association rule base indicating process failure → lack of responsibility, and uses each process failure node as a preceding term to match the corresponding subsequent type of lack of responsibility. For example, process failure node B1 matches rule C2 with a confidence level of 79.2% and a causal strength of 0.68, corresponding to inadequate supervision by management personnel (C2).

[0080] By retrieving the job descriptions for power grid operation and maintenance positions, the core responsibilities of the execution, management, and supervisory levels were clarified, and the responsible parties were further verified by combining this information with personnel operation records. Regarding the inspector's failure to inspect C1: The monthly inspection responsibility allocation record for the equipment was checked, confirming that inspector Zhang was the designated inspector for the equipment recently, and that the seal inspection items in his inspection record were all blank, with no on-site sign-in confirmation record, verifying that Zhang was the directly responsible party at the execution level; Regarding the inadequate supervision by management personnel (C2): An investigation of the process supervision records of Li, the team leader of the maintenance team, revealed that he failed to review the service life of the sealing components every quarter as required by the system, and also failed to urge the inspectors to rectify the blank inspection items. This verifies that Li is an indirect responsible party of the management. Based on the power grid safety supervision responsibility list, the Safety Supervision Department is required to conduct a special supervision of the operation and maintenance process every six months. The review found that the substation's seal replacement process had not been supervised recently. Matching rule B1→C3 was added, with a confidence level of 72.1% and a causal strength of 0.63, confirming a lack of supervision and inspection at the supervisory level, specifically C3. Personnel responsibility factors were determined: based on the rule matching and record verification results, the final responsibility deficiencies were identified as C1 - Inspector Zhang's missed inspection, C2 - Inadequate supervision by the operation and maintenance team leader Li, and C3 - Lack of supervision and inspection by the Safety Supervision Department. The responsible parties, types of responsibility, and legal basis for each node were clearly defined.

[0081] Finally, an inner-loop responsibility tracing mechanism is adopted. Based on the constructed logic tree, especially the sub-nodes of missing responsibility and process failure, reverse tracing is performed along the predefined three-level responsibility chain of execution layer → management layer → supervision layer. Using the preset responsibility boundaries of each level as a benchmark, the process failure sub-nodes are matched to the responsibility scope of the corresponding level, and then the specific responsible entity at that level is identified through the sub-nodes of missing responsibility, ultimately forming a complete location link of process failure sub-node → level responsibility matching → missing responsibility sub-node → specific responsible person.

[0082] Execution layer: The associated process failure sub-nodes are omissions in the inspection process (B2) and violations of basic operation procedures; the associated responsibility deficiency sub-nodes are omissions by inspectors (C1) and non-standard operations. Management: The related process failure sub-nodes are failure to implement the regularly changed process (B1), lack of planning and inadequate training process; the related responsibility failure sub-nodes are inadequate supervision by management personnel (C2) and failure to track the plan.

[0083] Supervisory layer: The associated process failure sub-nodes are all process failure sub-nodes, and the associated responsibility deficiency sub-nodes are the lack of supervision and inspection by the supervisory layer (C3) and the failure to hold people accountable in a timely manner.

[0084] The connection between each level and its sub-nodes is essentially a causal chain of responsibility performance status → process execution result → responsibility determination. The lack of responsibility at the execution and management levels directly leads to the generation of corresponding process failure sub-nodes. The lack of responsibility at the supervisory level allows process failure sub-nodes to continue to exist. The responsibility-deficient sub-nodes are a precise characterization of the lack of responsibility at each level, ultimately achieving the goal of tracing back to the matter, management, people, and responsibilities.

[0085] Finally, the fuzzy comprehensive evaluation method was used to determine the responsibility weights at each level. The specific steps are as follows: Determine the evaluation factor set: U = {Execution level responsibility U1, Management level responsibility U2, Supervisory level responsibility U3}; Determine the weight set: Based on expert scores (3 power grid safety management experts), the initial weights were calculated using the analytic hierarchy process (AHP), and after consistency verification, the weight set W = {0.4, 0.35, 0.25} was determined; Determine the comment set: V = {Main responsibility, Secondary responsibility, General responsibility}, with corresponding quantitative values ​​of {0.8, 0.5, 0.2}; Calculate the comprehensive evaluation result: Through fuzzy matrix operations, the evaluation result for execution level responsibility was 0.75 (main responsibility), for management level responsibility it was 0.6 (secondary responsibility), and for supervisory level responsibility it was 0.3 (general responsibility); The comprehensive evaluation result was calculated by combining the evaluation factor set U, the weight set W, and the comment set V using fuzzy matrix operations, as follows: For each element in the evaluation factor set U, determine its membership degree to the elements in each comment set. Take the average expert scores to form an n×m matrix R, where n is the number of elements in U and m is the number of elements in V, reflecting the degree of correlation between the two. Calculate B=[b1,b2,…,bm] using the formula B=W・R (matrix multiplication), representing the comprehensive membership degree of all elements in the evaluation factor set U to the elements in the comment set.

[0086] Since the elements in the evaluation factor set U in the power grid responsibility tracing have clear responsibility directions, the single-factor evaluation vector × V quantization value is used first (the element Ui in the evaluation factor set U corresponds to the row vector V quantization value of R); it can also be calculated through B·V quantization value. The final result matches the threshold of the comment set, and when the value is ≥ threshold 0.7, it is set as the main responsibility, and the evaluation is completed.

[0087] It is understood that the above system item embodiments correspond to the method item embodiments of the present invention, and can realize the power grid safety inspection and management method based on responsibility logic chain tracing and dual-ring collaboration provided by any of the above method item embodiments of the present invention.

[0088] It should be noted that the system embodiments described above are merely illustrative, and some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the system embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can be specifically implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0089] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A power grid safety inspection and management method based on responsibility logic chain tracing and dual-loop collaboration, characterized in that, include: Obtain equipment failure reports from faulty devices; Based on the equipment failure report, identify several defective links in the faulty equipment during the safety inspection process and generate rectification tasks for each defective link. For each rectification task, the executors, supervisors, and acceptance personnel are identified, and a task flow list consisting of several task nodes is created based on the executors, supervisors, and acceptance personnel; wherein, the task nodes include: execution nodes, supervision nodes, and acceptance nodes; Based on the task process list corresponding to each defective link, rectification tasks are issued to each of the executors, and the task completion status of each task node is monitored. When the task completion status of any task node is determined to be completed, the next task node is determined and a reminder is issued to the target personnel of the next task node; when the unchanged duration of the task completion status of any task node exceeds a preset time limit, an alarm message is issued to the target personnel of the task node until the task process list is completed; wherein, the target personnel include: executors, supervisors and acceptors.

2. The power grid safety inspection and management method based on responsibility logic chain tracing and dual-ring collaboration as described in claim 1, characterized in that, The step of identifying several defective links in the faulty equipment during the safety inspection process based on the equipment fault report and generating rectification tasks for each defective link includes: Obtain the operation and maintenance process plan for the faulty equipment; Based on the equipment failure report, a defect analysis is performed on each maintenance step in the maintenance process plan to identify several defective steps in the execution of the maintenance process plan and a description of the defects in each defective step. Based on each defective link and its description, a rectification task for each defective link is matched in a pre-set rectification task database. The rectification task database contains pre-set rectification tasks for several defects in each maintenance link.

3. The power grid safety inspection and management method based on responsibility logic chain tracing and dual-ring collaboration as described in claim 2, characterized in that, The step involves performing defect analysis on each maintenance stage of the maintenance process plan based on the equipment failure report, identifying several defective stages in the execution of the maintenance process plan, and describing the defects of each stage, including: Obtain the inspection work orders and maintenance work orders for the area where the faulty equipment is located; Perform fault analysis on the equipment fault reports to determine the fault characteristics of the faulty equipment; Based on the fault characteristics, inspection rules and maintenance rules are extracted from the operation and maintenance process plan. Based on the inspection rules and maintenance rules, execution defect analysis is performed on the inspection work order and the maintenance work order to determine several defective links and defect descriptions of each defective link.

4. The power grid safety inspection and management method based on responsibility logic chain tracing and dual-ring collaboration as described in claim 3, characterized in that, The fault characteristics include: several faulty parts; the inspection rules include: the planned inspection cycle of several faulty parts; the maintenance rules include: the planned maintenance time of several faulty parts. The process involves extracting inspection rules and maintenance rules from the operation and maintenance plan, performing execution defect analysis on the inspection work orders and maintenance work orders based on the inspection rules and maintenance rules, identifying several defective links and defect descriptions for each defective link, including: Based on the keywords extracted from the faulty parts in the operation and maintenance process plan, the planned inspection cycle and planned maintenance time of the faulty parts are determined. Based on the faulty part, several actual inspection cycles and actual inspection contents in the inspection work order, select the target inspection cycle related to the faulty part. The planned inspection cycle is matched with the target inspection cycle. When there is a planned inspection cycle that does not match the target inspection cycle, inspection defect information is generated based on the unmatched planned inspection cycle to characterize the failure of the faulty part to be inspected on time. Based on the preset time interval and the planned maintenance time, a record retrieval interval is generated for each faulty part. Based on the record retrieval interval, the maintenance record for each faulty part is retrieved in the maintenance work order. When no maintenance record for any part is found in the record retrieval interval, maintenance defect information is generated to characterize the failure to maintain the faulty part on time, based on the planned maintenance time when no maintenance record is found.

5. The power grid safety inspection and management method based on responsibility logic chain tracing and dual-ring collaboration as described in claim 4, characterized in that, After identifying several defective links and their defect descriptions, the process also includes: Obtain a preset responsibility tracing rule base; wherein, the responsibility tracing rule base stores several tracing association rules, and the tracing association rules are constructed by using an association rule mining algorithm to analyze the causal relationship strength between historical faulty equipment and each historical defective link based on historical fault analysis data; Based on the fault characteristics and the defective links, semantic similarity matching is performed in the responsibility tracing rule base to determine the target tracing association rules corresponding to each defective link; Based on the target tracing association rules, determine the causal relationship strength between each defective link and the faulty equipment. Based on the strength of the causal relationship, the responsibility ratio of each defective link is calculated, and the defective link with the highest responsibility ratio is identified as the primary defective link.

6. A power grid safety inspection and management system based on responsibility logic chain tracing and dual-loop collaboration, characterized in that, include: The report acquisition module is used to acquire equipment failure reports from faulty devices. The task generation module is used to determine several defective links of the faulty equipment during the safety inspection process based on the equipment fault report and generate rectification tasks for each defective link. The list creation module is used to determine the executors, supervisors, and acceptance personnel for each rectification task, and to create a task flow list consisting of several task nodes based on the executors, supervisors, and acceptance personnel respectively; wherein, the task nodes include: execution nodes, supervision nodes, and acceptance nodes. The task monitoring module is used to issue rectification tasks to each of the executors according to the task process list corresponding to each defective link, and to monitor the task completion status of each task node. The task alarm module is used to determine the next task node when the task completion status of any task node is determined to be completed, and to issue a reminder to the target personnel of the next task node; when the unchanging time of the task completion status of any task node exceeds a preset time limit, an alarm message is issued to the target personnel of the task node until the task process list is completed; wherein, the target personnel include: executors, supervisors and acceptance personnel.

7. The power grid safety inspection and management system based on responsibility logic chain tracing and dual-ring collaboration as described in claim 6, characterized in that, The task generation module, based on the equipment fault report, identifies several defective links in the faulty equipment during the safety inspection process and generates rectification tasks for each defective link, including: Obtain the operation and maintenance process plan for the faulty equipment; Based on the equipment failure report, a defect analysis is performed on each maintenance step in the maintenance process plan to identify several defective steps in the execution of the maintenance process plan and a description of the defects in each defective step. Based on each defective link and its description, a rectification task for each defective link is matched in a pre-set rectification task database. The rectification task database contains pre-set rectification tasks for several defects in each maintenance link.

8. The power grid safety inspection and management system based on responsibility logic chain tracing and dual-ring collaboration as described in claim 7, characterized in that, The task generation module, based on the equipment fault report, performs defect analysis on each maintenance step in the maintenance process plan, identifies several defective steps in the execution of the maintenance process plan, and describes the defects of each defective step, including: Obtain the inspection work orders and maintenance work orders for the area where the faulty equipment is located; Perform fault analysis on the equipment fault reports to determine the fault characteristics of the faulty equipment; Based on the fault characteristics, inspection rules and maintenance rules are extracted from the operation and maintenance process plan. Based on the inspection rules and maintenance rules, execution defect analysis is performed on the inspection work order and the maintenance work order to determine several defective links and defect descriptions of each defective link.

9. The power grid safety inspection and management system based on responsibility logic chain tracing and dual-ring collaboration as described in claim 8, characterized in that, The fault characteristics include: several faulty parts; the inspection rules include: the planned inspection cycle of several faulty parts; the maintenance rules include: the planned maintenance time of several faulty parts. The task generation module extracts inspection rules and maintenance rules from the operation and maintenance process plan, and performs execution defect analysis on the inspection work order and the maintenance work order according to the inspection rules and maintenance rules to determine several defective links and defect descriptions for each defective link, including: Based on the keywords extracted from the faulty parts in the operation and maintenance process plan, the planned inspection cycle and planned maintenance time of the faulty parts are determined. Based on the faulty part, several actual inspection cycles and actual inspection contents in the inspection work order, select the target inspection cycle related to the faulty part. The planned inspection cycle is matched with the target inspection cycle. When there is a planned inspection cycle that does not match the target inspection cycle, inspection defect information is generated based on the unmatched planned inspection cycle to characterize the failure of the faulty part to be inspected on time. Based on the preset time interval and the planned maintenance time, a record retrieval interval is generated for each faulty part. Based on the record retrieval interval, the maintenance record for each faulty part is retrieved in the maintenance work order. When no maintenance record for any part is found in the record retrieval interval, maintenance defect information is generated to characterize the failure to maintain the faulty part on time, based on the planned maintenance time when no maintenance record is found.

10. The power grid safety inspection and management system based on responsibility logic chain tracing and dual-ring collaboration as described in claim 9, characterized in that, Also includes: Responsibility tracing module; The responsibility tracing module is used to obtain a preset responsibility tracing rule base after determining several defective links and the defect descriptions of each defective link; wherein, the responsibility tracing rule base stores several tracing association rules, and the tracing association rules are constructed by using an association rule mining algorithm to analyze the causal relationship strength between historical faulty equipment and each historical defective link based on historical fault analysis data; Based on the fault characteristics and the defective links, semantic similarity matching is performed in the responsibility tracing rule base to determine the target tracing association rules corresponding to each defective link; Based on the target tracing association rules, determine the causal relationship strength between each defective link and the faulty equipment. Based on the strength of the causal relationship, the responsibility ratio of each defective link is calculated, and the defective link with the highest responsibility ratio is identified as the primary defective link.

Images