Intelligent power distribution equipment monitoring and fault early warning system and method

Abstract

The invention belongs to the technical field of power distribution equipment monitoring, and relates to an intelligent power distribution equipment monitoring and fault early warning system and method, and the system comprises an acquisition unit which is used for obtaining an electric parameter, an environment parameter and an operation state quantity of power distribution equipment and forming time synchronization feature data; the topology unit is used for generating a local topology model according to the relation between the cabinet, the loop and the busbar and outputting topology attributes; the evaluation unit is used for generating an operation state evaluation result based on the feature data and the topological attributes and updating an early warning threshold value; the governance unit is used for performing deduplication, suppression and credibility calculation on the alarm to form a disposal level; the interlocking control unit is used for checking and outputting a control action when the disposal level meets an interlocking condition; and the operation and maintenance unit is used for generating a preventive test plan and updating parameters based on a test result. According to the technical scheme of the invention, topological association and threshold update are introduced into monitoring, evaluation, disposal and control links, so that state recognition and control execution are kept consistent.

Description

Technical Field

[0001] This invention belongs to the field of power distribution equipment monitoring technology, and specifically relates to an intelligent power distribution equipment monitoring and fault early warning system and method. Background Technology

[0002] In existing technologies, intelligent power distribution systems typically rely on power distribution monitoring devices and remote management platforms to display the operating status of power distribution equipment by collecting electrical parameters, environmental parameters, and operational status parameters, thereby meeting daily operation and maintenance needs. However, existing systems generally use fixed thresholds and direct alarm triggering methods. When load levels change with daily and weekly cycles, and when there are differences in equipment aging and heat dissipation conditions, the applicability of fixed thresholds is not constant at different operating stages, and the results of status assessment feedback fluctuate over time.

[0003] Existing systems typically send alarm events one by one in chronological order. However, when there are topological relationships in the power distribution network, operational anomalies may propagate in the topology, causing alarms to appear in a concentrated manner within a short period of time. The correlation and scope of impact between alarms are not easily presented, increasing the workload of maintenance personnel.

[0004] Meanwhile, existing systems mostly rely on alarms to trigger remote switching actions, which are insufficient in verifying personnel protective clothing, access control status, and area occupancy status, resulting in instability in the correspondence between control actions and on-site status.

[0005] In addition, existing operation and maintenance methods rely heavily on experience in organizing inspection projects and maintenance tasks, resulting in a discontinuous relationship between changes in equipment status and maintenance schedules.

[0006] It is evident that existing technologies often suffer from insufficient adaptability in state assessment, delays in alarm handling response, and unstable consistency. These are the shortcomings of existing technologies.

[0007] In view of this, it is very necessary to provide an intelligent power distribution equipment monitoring and fault early warning system and method to solve the above-mentioned defects in the prior art. Summary of the Invention

[0008] The purpose of this invention is to address the shortcomings of the existing technology, such as insufficient adaptability of state assessment, delayed alarm handling response, and unstable consistency performance, by providing a smart power distribution equipment monitoring and fault early warning system and method to solve the above-mentioned technical problems.

[0009] To achieve the above objectives, the present invention provides the following technical solution: A smart power distribution equipment monitoring and fault early warning system includes: The data acquisition unit is used to acquire electrical parameters, environmental parameters, and equipment operating status data of the power distribution equipment, and to generate time-synchronized characteristic data. Topology units are used to generate a local topology model of the power distribution room based on the relationships between the cabinets, circuits and busbars of the power distribution equipment, and output the topology attributes corresponding to the feature data. The evaluation unit is used to generate evaluation results of the operating status based on feature data and topology attributes, and to update the early warning thresholds based on the operating history and load pattern, on the basis of the preset early warning thresholds. The governance unit is used to deduplicatize, suppress, and calculate the credibility of alarms generated from the assessment results, and to form a handling level. The interlocking control unit is used to verify the remote control command when the handling level meets the interlocking conditions, and to output control actions according to the safety interlocking strategy under the premise of ensuring personal and area safety; The operation and maintenance unit is used to generate preventive test plans and maintenance task arrangements based on topology attributes and handling levels, and to update warning thresholds, handling levels and security interlocking strategies based on test results.

[0010] By adopting the above technical solution, topology recognition, adaptive state assessment and interlocking execution logic are introduced into the process of monitoring, representation, evaluation, handling and control, continuous identification and consistent response to the operating status of power distribution equipment can be achieved. It can maintain the stability of the state assessment performance under the conditions of load changes, environmental differences and equipment operating years, and meet the requirements of alarm presentation correlation, control action correspondence with field status and maintenance rhythm continuity during operation.

[0011] The data acquisition unit obtains operational data to form a monitoring foundation with time-series characteristics; the topology unit generates a local topology model that enables the representation of operational status to have structural correlation information; the assessment unit introduces factors of operational history and load form on the baseline early warning conditions to ensure that the status feedback remains adaptable across different operational segments; the management unit performs correlation analysis and consistency processing on alarms to ensure that alarms are merging and directional; the interlocking control unit verifies the remote control actions in conjunction with relevant on-site conditions to ensure that the implementation of control commands corresponds to the on-site status; the operation and maintenance unit organizes test and maintenance arrangements based on topology attributes and handling levels to ensure that changes in equipment operating status and maintenance rhythm form a continuously trackable relationship; through the synergistic effect of these functions, the status assessment and handling execution process remain corresponding in both time and structural dimensions, the operational feedback is stable, and the operation and maintenance organization information is more traceable and interpretable, meeting the requirements of power distribution operation and maintenance for the adaptability of status assessment, alarm correlation presentation, and safety consistency of control actions.

[0012] Preferably, when generating a local topology model, the topology unit represents the local topology model in a structured way with nodes and connection relationships and generates topology attributes. The topology attributes include the influence radius of each node relative to the target object. The influence radius is used to characterize the transmission level in the local topology model when an anomaly propagates from any node to the target object. The topology unit generates an influence radius sorting based on the influence radius, which is used to indicate the propagation order and propagation range of nodes.

[0013] By forming a representable structural system of the relationships between power distribution equipment and introducing a quantitative representation to describe the transmission hierarchy in this topological unit, the following technical effects can be achieved: First, the presentation of operational status has changed from an independent display of single points to a structured expression with correlations. This allows operational changes caused by multiple device nodes to show the order and direction of their impact in the network structure. This enables maintenance personnel to obtain structurally oriented reference information when anomalies occur, transforming the presentation of information from planar to hierarchical and improving the efficiency of location. Second, by utilizing the quantitative representation of hierarchical transmission relationships, a distinguishable propagation link can be formed when there are multiple possible alarm sources. This allows the operational deviation caused by equipment aging, load transfer, or wiring relationship adjustment to reflect different action chain levels in the topology. As a result, the associated anomalies no longer appear as disordered superpositions, but rather have a deducible sequential relationship, thereby enhancing the clarity of the status interpretation. Third, the hierarchical ordering can provide an executable priority for subsequent operational processing strategies, so that operational handling no longer depends on experience judgment or fixed rules, but forms a verifiable execution path in the topology, making the connection between operation and maintenance actions logically consistent, and improving the consistency of maintenance decisions while ensuring operational security.

[0014] As a preferred method, the warning threshold is updated based on the operating history and load pattern, including: performing quantile modeling on the feature data based on the daily and weekly load patterns to generate candidate warning thresholds; verifying the candidate warning thresholds by playing back historical data; updating the warning threshold with the candidate warning thresholds and generating corresponding warning threshold change entries when the verification passes; and keeping the warning threshold unchanged and generating warning threshold retention entries when the verification fails.

[0015] This assessment unit incorporates the periodic characteristics of historical operating records and load patterns into the generation and adjustment of early warning thresholds, and uses quantile modeling and playback verification as the basis for judging threshold changes, achieving the following technical effects: First, it ensures that the warning threshold remains adaptable in different operating phases. During operation, there are periodic changes such as day-night load switching and weekdays versus non-weekdays. When using a fixed threshold, misjudgments are likely to occur. This mechanism enables the threshold to be adjusted in an interpretable and gradual manner according to changes in operating behavior patterns, so that the status judgment shows consistency in different operating segments. Second, by placing the threshold adjustment process in the historical data playback test, the threshold update does not depend on a single fluctuation or short-term disturbance, but is verified based on a representative operating sequence, making the threshold update judgment verifiable and traceable, and making the early warning and alarm behavior continuous and stable in the time dimension. Third, by generating entries for each threshold change, the threshold change process becomes a traceable record, enabling subsequent maintenance personnel to identify the triggering conditions, update segments, and execution background of the threshold change from the record. This allows the change of the warning threshold to form an interpretable path, ensuring that the status assessment has a consistent and reproducible performance in operation management, decision auditing, and maintenance organization.

[0016] As a preferred method, the alarms generated by the evaluation results are deduplicated and suppressed, including: using at least one of the following: a similar alarm merging mechanism, a backoff suppression mechanism, and a root cause merging mechanism. The similar alarm merging mechanism merges alarms from the same source or the same topological domain within a time window. The backoff suppression mechanism sets a cooling-off period for similar alarms within a time window. The root cause merging mechanism merges alarms derived from upstream key nodes into a single root cause entry based on the upstream key nodes within the influence radius.

[0017] This governance unit aggregates and merges generated alarm information by utilizing alarm processing logic with temporal and topological correlations, and formulates a controllable suppression strategy for repeatedly triggered alarms, achieving the following technical effects: First, multiple alarms caused by the same operational phenomenon are no longer presented as independent entries, but rather as a unified alarm result that can be merged. This prevents alarm outputs from being concentrated and stacked indiscriminately in a short period of time, allowing maintenance personnel to directly identify the dominant alarm source in the interface and reduce the burden of manually filtering through a large number of repetitive prompts. Second, by setting cooling and suppression conditions for alarms that are frequently triggered in a short period of time, the instantaneous disturbances, rapid fluctuations or short-term load switching in the operating data will no longer cause continuous alarm triggering. This allows the system to maintain the stability and continuity of alarm output in scenarios with large signal fluctuations or frequent switching of operating conditions, making the alarm behavior more closely reflect the actual changes in the long-term state. Third, by identifying key nodes upstream of the propagation link based on topological hierarchical relationships and merging related downstream derived alarms, the system can identify source alarms based on structural relationships, making the presentation of anomalies directional and traceable. This allows maintenance responses to no longer process downstream multi-point manifestations one by one, but to directly lock onto the source device with dominant influence, transforming the location path from multi-node search to structured directional query, thus achieving stable performance in fault location efficiency and the accuracy of handling decisions.

[0018] Preferably, the credibility calculation includes: calculating the credibility score of the alarm based on the weight parameters of multi-source measurement consistency, topology consistency and sensor health. Multi-source measurement consistency is used to characterize the degree of correlation of multi-channel measurement data in the time domain, topology consistency is used to characterize the propagation correlation of alarm in topology attributes, and sensor health is used to characterize the working stability of the acquisition source. The weight parameters are generated by system configuration or learning process, and the credibility score is related to the handling level.

[0019] This governance unit achieves the following technical effects by incorporating a comprehensive measurement model that considers multi-source measurement consistency, topological consistency, and the health status of the data acquisition sources into the generated alarms, and forming a weighted reliability score: First, it enables alarm results to reflect the mutual corroboration relationship of multi-channel data in the time domain. When multiple measurement channels show a consistent trend for the same operating phenomenon, the reliability score increases. When there are differences or abnormal deviations between signals, the reliability score decreases. This makes alarms no longer based solely on single-point triggering, but forms a stable and reliable evaluation in a time-related information environment, thus making alarm performance closer to the actual operating state. Second, by placing alarms in the topology attributes for correlation analysis of propagation relationships, alarms no longer exist as isolated events, but have structural traceability. When there is a stable propagation link between an alarm and other nodes, the credibility score can reflect the position and correlation of the alarm in the topology structure, enabling the system to identify structural anomalies and making the state interpretation more directional and hierarchical, thereby improving the effectiveness of location information in operation and maintenance response. Third, the alarm results are verified by combining the health status of the data acquisition source, so that the stability of sensor operation and measurement reliability are reflected in the evaluation. When there is offset, drift or data jitter in the data acquisition source, the reliability score can be adaptively lowered to avoid false alarms caused by abnormal acquisition. This ensures that the alarm output can reflect the real working condition and forms an endogenous protection mechanism for data quality. This enables the system to maintain the consistency and reliability of the operating status assessment in long-term operation, and makes the generation process of the handling level interpretable and traceable.

[0020] As a preferred option, the interlocking conditions are defined by the interlocking strategy set, which includes the protective equipment status, access control status, area occupancy status, and operation confirmation status. It can be configured according to the operating scenario. When the interlocking conditions are not met, the interlocking control unit outputs a rejection message or a review request and records an audit log.

[0021] By introducing a multi-condition verification system related to the field status before the control command is executed and recording the interlocking execution process information, the following technical effects can be achieved in this interlocking control unit: First, it enables remote control actions to be correlated with on-site work behavior. When there is personnel entry, equipment operation, or work preparation, the interlocking strategy set can intercept or delay confirmation of control commands, so that the triggering of control actions is correlated with the actual work status, and the remote scheduling behavior has a consistent execution basis during operation. Second, by combining the determination of the status of protective clothing, access control and area occupancy, the control decision can reflect the dynamic changes of the environment in which the equipment is located. When there are potential risks on site or safety preparations are not completed, the interlocking conditions can form a clear interception signal, enabling remote operation to have the necessary action screening capabilities in scenarios where attention to personal risks is required, and making the control logic adaptable and continuous in different operation and maintenance scenarios. Third, by logging rejection information or review requests, the execution process of control actions has a traceable basis. Operation and maintenance personnel can identify the time, background and condition changes of interlock triggering from the records, so that remote control behavior has a clear traceability path in operation management, division of responsibilities and subsequent evaluation. This enables the system to maintain consistent control behavior with the field status during long-term continuous operation and supports the management side to dynamically adjust operation and maintenance strategies.

[0022] As a preferred option, the interlocking control unit executes a local autonomous strategy when communication is abnormal. The local autonomous strategy includes: determining alarms based on cached warning thresholds and handling levels, limiting whitelist control actions and generating event logs; after communication is restored, reconciling the event logs during the autonomous period with cloud records and performing playback verification.

[0023] In the case of communication failure, the interlocking control unit executes a local autonomous strategy. By using cached information during autonomous operation to maintain the executability of alarm judgment and control actions, and performing consistency reconciliation and playback verification after communication is restored, the following technical effects can be achieved: First, it enables the system to maintain basic operational capabilities under conditions of communication link interruption, delay, or fluctuation. The autonomous strategy uses the cached warning threshold and handling level as the basis for judgment, so that alarm judgment does not depend on the real-time uplink, the feedback of the operating status is not interrupted, and the equipment can still maintain continuous operation even without cloud support. Second, by limiting whitelist control actions, the control behavior during the autonomous period is kept within a verified and safe execution range, avoiding control commands that do not meet the operating conditions due to control link mismatch caused by communication anomalies. This makes the execution under the autonomous strategy predictable and safe and controllable, and enables the operating organization to have edge stability in the remote disconnection state. Third, after communication is restored, consistency reconciliation and playback verification of the event logs and cloud records formed during the autonomous period are performed to ensure that the operational behavior, alarm changes and control execution during the autonomous period form an auditable complete link. This enables the system to confirm the consistency between autonomous behavior and normal operation behavior, and to continuously express the trend of state changes in the operation records. This provides reliable support for the consistency, traceability and executive execution of the control process during long-term system operation.

[0024] Preferably, a preventive test plan is generated based on topological attributes and treatment levels, including: selecting target objects that are in the order of influence radius based on the influence radius ranking and treatment level, and generating test execution paths for the preventive test plan based on the interdependencies between target objects. The test execution paths identify the inspection order, adjustment conditions, and priority of the target objects. The preventive test plan includes at least one of the following: target object identification, inspection items, execution order, and completion status.

[0025] This operation and maintenance unit generates preventative test plans with execution order based on topology attributes and treatment levels, and selects test objects and constructs paths based on influence radius ranking, achieving the following technical effects: First, it enables the generation of test plans to no longer rely on experience-based judgment or static periodic arrangements, but to be dynamically organized based on the structural relationships and impact range of the current operating status. This makes the selection of test objects interpretable and allows operation and maintenance activities to be carried out around nodes that may have a chain reaction effect, thereby forming a continuous relationship between test arrangements and operating status. Second, by constructing an experimental execution path that includes the testing sequence and adjustment conditions, the experimental behavior is no longer a discrete action executed independently, but has a logical connection between nodes based on the dependency relationship. When the state of a target object changes, the execution path can be adjusted according to the set conditions, so that the experimental process can maintain the integrity and consistency of the execution chain in different operating scenarios. Third, by including information such as target object identification, inspection items, execution sequence, and completion status in the test plan, the execution process of preventive tests is recordable and traceable. Maintenance personnel can clearly identify the test progress from the plan and execution records, and the maintenance behavior has a clear record link at the management and audit level. This ensures a consistent correspondence between equipment status changes and test execution records in long-term operation and maintenance, and makes the organization and results of operation and maintenance stable and continuous.

[0026] Preferably, the acquisition unit performs time synchronization and data quality marking. The data quality marking is used to indicate the anomaly type, which includes at least one of out-of-bounds, mutation, and packet loss. When generating the evaluation result, the evaluation unit performs interpolation or elimination strategies based on the data quality marking and records the processing result as a data quality entry.

[0027] By synchronizing the data acquisition time and marking the data quality in the acquisition and evaluation units, and processing and recording the processing results based on the data quality markings during the status evaluation, the following technical effects can be achieved: First, it ensures that data from different sources, sampling periods, and channels have a unified time reference before entering the evaluation stage, so that the analysis of operational status is no longer affected by time alignment errors, that time-series relationships can play a role in judging changing trends, and that the presentation of operational status has continuity and comparability. Second, by distinguishing different types of anomalies such as out-of-bounds, mutations, and packet loss in the data quality tags, the system can adopt differentiated strategies based on the nature of the anomalies during state assessment. Out-of-bounds data can be used as state features for analysis, mutation data can be smoothed or imputed, and packet loss data can be removed or reconstructed. This makes data processing no longer dependent on a single filtering logic, so that the assessment results can reflect the real trend of operational changes. Third, by recording the processed results as data quality items during the evaluation process, the operational status judgment process becomes traceable, enabling operation and maintenance personnel to identify data anomalies and their handling methods from the records, making the correspondence between data sources, processing strategies and evaluation conclusions verifiable, ensuring consistency and interpretability of operational status analysis in long-term operation, and supporting subsequent model correction or threshold adjustment strategies.

[0028] Furthermore, the present invention also provides a method for monitoring and fault early warning of intelligent power distribution equipment, comprising the following steps: Acquire electrical parameters, environmental parameters, and equipment operating status parameters of power distribution equipment, and generate time-synchronized characteristic data; A local topology model of the power distribution room is generated based on the relationship between the cabinets, circuits and busbars of the power distribution equipment, and the topology attributes corresponding to the feature data are output. The evaluation results of the operating status are generated based on feature data and topology attributes, and the warning thresholds are updated according to the operating history and load pattern based on the preset warning thresholds. Alarms are generated based on the assessment results, and deduplication, suppression, and confidence calculation are performed on the alarms to obtain the handling level. When the handling level meets the interlocking conditions, the remote control command is verified, and control actions are output according to the safety interlocking strategy under the premise of ensuring personal and area safety. Preventive test plans and maintenance task arrangements are generated based on topology attributes and response levels, and warning thresholds, response levels, and safety interlock strategies are updated based on test results.

[0029] By adopting the above technical solution, a coherent response relationship is formed between status identification, alarm handling, control execution and maintenance adjustment within the system, realizing a closed expression of the operating status from monitoring to execution and then to backtracking. This enables the operating information to maintain stable indication characteristics under conditions of load change, environmental change and equipment aging, and ensures that control actions are consistent with the field status.

[0030] Among them, the state expression composed of feature data and structural attributes provides relevant input for subsequent judgment; the introduction of operating history and load mode enables early warning judgment to have the ability to self-adjust over time; alarms are processed through correlation aggregation before entering the handling link to make the presented results have a centralized orientation characteristic; control commands are verified by on-site status before execution to make the action itself verifiable; maintenance activities are organized based on state indications and structural relationships to enable the maintenance rhythm to follow and adjust with changes in operating status; through the operation of the above-mentioned correlation chain, the power distribution system presents a continuous, reviewable and deducible state change path during operation.

[0031] The beneficial effects of this invention are that by introducing topology recognition, adaptive state assessment and interlocking execution logic into the process of monitoring, representation, evaluation, handling and control, it can achieve continuous identification and consistent response to the operating status of power distribution equipment. It can maintain the stability of the state assessment performance under the conditions of load changes, environmental differences and equipment operating years, and meet the requirements of alarm presentation correlation, control action correspondence with field status and maintenance rhythm continuity during operation.

[0032] Furthermore, the design principle of this invention is reliable, the structure is simple, and it has a very wide range of application prospects.

[0033] Therefore, it is evident that the present invention has outstanding substantive features and significant progress compared with the prior art, and the beneficial effects of its implementation are also obvious. Attached Figure Description

[0034] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0035] Figure 1 This is a schematic diagram of a smart power distribution equipment monitoring and fault early warning system provided by the present invention; Figure 2 This is a flowchart of a method for monitoring and fault early warning of intelligent power distribution equipment provided by the present invention.

[0036] The system comprises: 1. Acquisition unit; 2. Topology unit; 3. Evaluation unit; 4. Governance unit; 5. Interlocking control unit; and 6. Operation and maintenance unit. Detailed Implementation

[0037] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. The following embodiments are explanations of the present invention, but the present invention is not limited to the following implementation methods.

[0038] Example 1: like Figure 1 As shown in the figure, this embodiment provides an intelligent power distribution equipment monitoring and fault early warning system, including: The acquisition unit 1 is used to acquire electrical parameters, environmental parameters, and equipment operating status parameters of the power distribution equipment, and to generate time-synchronized characteristic data. Topology Unit 2 is used to generate a local topology model of the power distribution room based on the relationship between the cabinets, circuits and busbars of the power distribution equipment, and output the topology attributes corresponding to the feature data. Evaluation unit 3 is used to generate evaluation results of the operating status based on feature data and topology attributes, and to update the warning threshold based on the operating history and load pattern on the basis of the preset warning threshold. Governance Unit 4 is used to deduplicatize, suppress, and calculate the credibility of alarms generated from the assessment results, and to form a handling level; The interlocking control unit 5 is used to verify the remote control command when the handling level meets the interlocking conditions, and to output control actions according to the safety interlocking strategy under the premise of ensuring personal and area safety; Operation and maintenance unit 6 is used to generate preventive test plans and maintenance task arrangements based on topology attributes and handling levels, and update early warning thresholds, handling levels and security interlocking strategies based on test results.

[0039] By adopting the above technical solution, topology recognition, adaptive state assessment and interlocking execution logic are introduced into the process of monitoring, representation, evaluation, handling and control, continuous identification and consistent response to the operating status of power distribution equipment can be achieved. It can maintain the stability of the state assessment performance under the conditions of load changes, environmental differences and equipment operating years, and meet the requirements of alarm presentation correlation, control action correspondence with field status and maintenance rhythm continuity during operation.

[0040] Specifically, the operational data acquired by the acquisition unit 1 is used to form a monitoring foundation with time-series characteristics; the local topology model generated by the topology unit 2 enables the representation of operational status to have structural correlation information; the evaluation unit 3 introduces factors of operational history and load form on the baseline early warning conditions to ensure that the status feedback remains adaptable across different operational segments; the management unit 4 performs correlation analysis and consistency processing on alarms to ensure that alarms exhibit merging and directional characteristics; the interlocking control unit 5 verifies the relevant on-site conditions before executing remote control actions to ensure that the implementation of control commands corresponds to the on-site status; the operation and maintenance unit 6 organizes test and maintenance arrangements based on topology attributes and handling levels to ensure that changes in equipment operating status and maintenance rhythm form a continuously trackable relationship; through the synergistic effect of these functions, the status assessment and handling execution process maintain correspondence in both time and structure dimensions, the operational feedback is stable, and the operation and maintenance organization information is more traceable and interpretable, meeting the requirements of power distribution operation and maintenance for the adaptability of status assessment, alarm correlation presentation, and safety consistency of control actions.

[0041] Hereinafter, based on embodiments of this application, reference is made to... Figure 1 The schematic diagram shown above provides a detailed explanation of each unit in the system.

[0042] In this embodiment of the application, the acquisition unit 1 is set on the field side of the power distribution equipment to acquire the electrical parameters, environmental parameters and equipment operating status of the power distribution equipment, and form time-synchronized feature data; the acquisition unit 1 may include an acquisition interface for accessing various sensing devices, a time reference and timestamp processing structure, and a data processing structure for aligning data and constructing features.

[0043] Specifically, the acquisition interface is used to obtain raw measurement data from power distribution equipment and its associated measurement points. Electrical parameters include at least one or more of the following: voltage, current, active power, reactive power, power factor, three-phase imbalance, and harmonic parameters; environmental parameters include at least one or more of the following: temperature, humidity, water immersion, smoke, and cabinet door magnetic field; equipment operating status parameters include at least one or more of the following: circuit breaker opening / closing position, relay protection device alarm status, controller operating status, and manual operation records. Simultaneously, acquisition unit 1 establishes separate data channels for raw quantities from different sources, independently completing data access and buffering.

[0044] Furthermore, the acquisition unit 1 is equipped with a unified time reference and timestamp generation structure to establish a unified time domain among different quantity classes. A timestamp is appended to each raw data record to characterize the sampling time. Considering the difference in sampling frequencies between electrical parameters and environmental parameters, the acquisition unit 1 can use window alignment to establish a correspondence between different channels within the same time window. If a channel does not obtain valid data within the window, the acquisition unit 1 retains a missing data marker and does not perform interpolation replacement to maintain the temporal integrity of the raw quantities.

[0045] After window alignment is completed, acquisition unit 1 can add data quality tags to the data records to indicate anomaly types, such as out-of-bounds values, abrupt changes, or missing values, so that subsequent units can distinguish between valid data and data to be verified when using it. For example, the data quality tags are stored synchronously with the original data records.

[0046] By synchronizing the collected data with time and marking the data quality, the acquisition unit 1 can achieve the following technical effects: First, it ensures that data from different sources, sampling periods, and channels have a unified time reference before entering the evaluation stage, so that the analysis of operational status is no longer affected by time alignment errors, that time-series relationships can play a role in judging changing trends, and that the presentation of operational status has continuity and comparability. Second, by distinguishing different types of anomalies such as out-of-bounds, mutations, and packet loss in the data quality tags, the system can adopt differentiated strategies based on the nature of the anomalies during state assessment. Out-of-bounds data can be used as state features in the analysis, mutation data can be smoothed or imputed, and packet loss data can be removed or reconstructed. This makes data processing no longer dependent on a single filtering logic, so that the assessment results can reflect the real trend of operational changes.

[0047] In this embodiment, the acquisition unit 1 can construct feature data based on the aligned data, including a standardized expression of the original quantity and derived quantities derived from the original quantity. The derived quantities can be obtained by converting the proportional relationship between the three-phase quantities, the phase difference, the relationship between current and power, etc. For example, both the feature data and the original data are associated with the device identifier and circuit identifier of the acquisition point, so that they can be mapped to the specific location of the power distribution system in the upper-level processing.

[0048] Furthermore, acquisition unit 1 supports local caching to ensure data availability under communication-constrained conditions. Upon communication recovery, acquisition unit 1 re-transmits the previously unuploaded feature data in chronological order to maintain the integrity of the upload sequence. Acquisition unit 1 provides a unified access interface to the upper-layer system, which supports reading feature data and its data quality markers by time interval, measurement point, or quantity category.

[0049] For example, the electrical parameter acquisition device can be set at the busbar of the circuit output terminal of the distribution cabinet, the temperature and humidity sensor can be set inside the cabinet, the smoke and water immersion detection device can be set at the bottom area of ​​the cabinet, and the door magnetic device can be set at the cabinet door; the electrical parameters can be acquired in sub-second cycles, the environmental parameters can be acquired in second cycles, and the data are uniformly output as feature data after forming window alignment.

[0050] In this embodiment of the application, the topology unit 2 can be set between the acquisition unit 1 and the evaluation unit 3, and is used to generate a local topology model of the power distribution room based on the relationship between the cabinets, circuits and busbars of the power distribution equipment, and output the topology attributes corresponding to the feature data; the topology unit 2 associates the feature data provided by the acquisition unit 1 with equipment identification, circuit identification and busbar affiliation, so that the upper-level processing can identify the structural position of the operating status in the spatial and electrical connection relationship of the power distribution room.

[0051] Specifically, topology unit 2 first establishes a set of equipment nodes based on the power distribution equipment layout information, including distribution cabinet nodes, loop nodes, and busbar nodes. Distribution cabinet nodes represent physical units within the cabinet, loop nodes represent outgoing loops within the cabinet, and busbar nodes represent the busbars to which the loops belong. Topology unit 2 can determine the connection relationships between nodes based on the equipment layout records, loop number records, and switchgear configuration records within the power distribution room. This includes the membership relationship between loop nodes and their respective distribution cabinet nodes, as well as the electrical connection relationship between loop nodes and their corresponding busbar nodes. The nodes and their relationships are then constructed into a structured topology representation to depict the internal electrical connection structure of the power distribution room.

[0052] In this embodiment, after generating a local topology model, topology unit 2 can associate feature data with the topology model. Since the feature data contains device identifiers and loop identifiers corresponding to the measurement points, topology unit 2 can locate the target node in the topology model based on these identifiers and attach topology attributes corresponding to the node to the feature data, including the device node's hierarchical position in the topology model, its busbar information, and its connection information with adjacent nodes. The introduction of topology attributes enables the differentiation and processing of the same type of measurement quantities at different physical locations.

[0053] In some embodiments of this application, when generating a local topology model, topology unit 2 can structurally represent the local topology model using nodes and connection relationships, and generate an influence radius based on the structured model. This radius characterizes the number of node levels required for any node to transmit its influence to the target node when an abnormality occurs. Specifically, when calculating the influence radius, topology unit 2 performs a graph traversal of all nodes in the model, centered on the target node, records the minimum connection level distance from each node to the target node, uses this distance as the influence radius value, and forms an influence radius sort to indicate the propagation order and range between nodes. This allows the upper-level processing to determine the transmission relationship of abnormal signals within the power distribution room based on the topological distance between nodes.

[0054] Topology attributes and influence radius ranking are used to interpret operational state variables during subsequent evaluation. For example, when a loop node changes, topology attributes enable the upper layer to identify the potential associated impacts of this change on other loops under the same busbar; when operating pressure or temperature rise occurs at a busbar node, influence radius ranking can indicate the set of potentially affected lower-level loop nodes, thus enabling the processing logic to distinguish between local anomalies and propagating anomalies with a wider impact range.

[0055] By forming a representable structural system of the relationships between power distribution equipment and introducing a quantitative representation to describe the transmission hierarchy in this topology unit 2, the following technical effects can be achieved: First, the presentation of operational status has changed from an independent display of single points to a structured expression with correlations. This allows operational changes caused by multiple device nodes to show the order and direction of their impact in the network structure. This enables maintenance personnel to obtain structurally oriented reference information when anomalies occur, transforming the presentation of information from planar to hierarchical and improving the efficiency of location. Second, by utilizing the quantitative representation of hierarchical transmission relationships, a distinguishable propagation link can be formed when there are multiple possible alarm sources. This allows the operational deviation caused by equipment aging, load transfer, or wiring relationship adjustment to reflect different action chain levels in the topology. As a result, the associated anomalies no longer appear as disordered superpositions, but rather have a deducible sequential relationship, thereby enhancing the clarity of the status interpretation. Third, the hierarchical ordering can provide an executable priority for subsequent operational processing strategies, so that operational handling no longer depends on experience judgment or fixed rules, but forms a verifiable execution path in the topology, making the connection between operation and maintenance actions logically consistent, and improving the consistency of maintenance decisions while ensuring operational security.

[0056] Furthermore, Topology Unit 2 supports updates to the local topology model. When the configuration of equipment, circuit allocation, or busbar structure within the power distribution room is adjusted, Topology Unit 2 can reconstruct the topology model based on the updated equipment list and connection records. After reconstruction, it updates the topology attributes and influence radius sorting in the same way. At the same time, it uses version identifiers to manage different batches of topology models, enabling the upper layer to call the corresponding model version when processing historical and real-time data, ensuring structural consistency.

[0057] For example, a distribution cabinet in a power distribution room includes multiple outgoing circuits, each belonging to the same busbar; when generating the model, topology unit 2 treats the distribution cabinet as a node, each outgoing circuit of the distribution cabinet as several circuit nodes, and the busbar as the superior node; the circuit nodes and the distribution cabinet nodes form a subordinate relationship, and the circuit nodes and the busbar nodes form an electrical connection relationship; the influence radius can be obtained from the hierarchical distance from the node to the busbar node to determine the topological relative position between the circuits, which is used to determine the propagation relationship of subsequent changes in operating status.

[0058] In this embodiment, the evaluation unit 3 is used to generate an evaluation result of the operating status based on feature data and topology attributes, and to update the warning threshold according to the operating history and load pattern based on a preset warning threshold. During this process, the evaluation unit 3 needs to process the feature data from the acquisition unit 1 and combine it with the topology attributes provided by the topology unit 2 to generate an evaluation result that reflects the current operating status, enabling subsequent processing to execute corresponding response logic according to different operating states.

[0059] Specifically, evaluation unit 3 maintains the matching relationship between operating status quantities and warning thresholds. Warning thresholds may include multiple sets of thresholds for electrical parameters, environmental parameters, and equipment operating status quantities, each set corresponding to different operating objects. Evaluation unit 3 compares the original and derived quantities in the feature data with the warning threshold sets to determine whether each feature quantity is in the normal or warning range under the current state. Topological attributes are used to establish the correspondence between feature quantities and equipment structure, enabling evaluation unit 3 to determine whether an anomaly is localized or structurally related.

[0060] In this embodiment, when updating the warning threshold, the evaluation unit 3 adjusts the warning threshold based on the operating history and load pattern. The operating history includes the recorded characteristic data sequence and its corresponding evaluation result sequence. The load pattern includes daily and weekly patterns. The evaluation unit 3 performs quantile modeling on the characteristic data based on the operating data within different time periods to obtain candidate warning thresholds. The quantile modeling is used to establish the statistical distribution of characteristic quantities under different load patterns, enabling the candidate warning thresholds to reflect changes in operating characteristics.

[0061] After generating candidate early warning thresholds, evaluation unit 3 verifies the candidate early warning thresholds through historical data playback. The playback process includes applying the candidate early warning thresholds to historical feature data sequences and comparing the generated evaluation results with recorded operational events to determine whether the candidate early warning thresholds conform to the known operational patterns of the system. When the verification results meet the consistency condition, evaluation unit 3 updates the current early warning threshold with the candidate early warning thresholds and generates an early warning threshold change entry. When the verification results do not meet the consistency condition, evaluation unit 3 keeps the early warning threshold unchanged and generates an early warning threshold hold entry. The early warning threshold entry records the basis, time point, and corresponding load pattern for threshold adjustment, for subsequent processing.

[0062] By incorporating the periodic characteristics of historical operation records and load patterns into the generation and adjustment of early warning thresholds, and using quantile modeling and replay verification as the basis for judging threshold changes, the following technical effects can be achieved in this evaluation unit 3: First, it ensures that the warning threshold remains adaptable in different operating phases. During operation, there are periodic changes such as day-night load switching and weekdays versus non-weekdays. When using a fixed threshold, misjudgments are likely to occur. This mechanism enables the threshold to be adjusted in an interpretable and gradual manner according to changes in operating behavior patterns, so that the status judgment shows consistency in different operating segments. Second, by placing the threshold adjustment process in the historical data playback test, the threshold update does not depend on a single fluctuation or short-term disturbance, but is verified based on a representative operating sequence, making the threshold update judgment verifiable and traceable, and making the early warning and alarm behavior continuous and stable in the time dimension. Third, by generating entries for each threshold change, the threshold change process becomes a traceable record, enabling subsequent maintenance personnel to identify the triggering conditions, update segments, and execution background of the threshold change from the record. This allows the change of the warning threshold to form an interpretable path, ensuring that the status assessment has a consistent and reproducible performance in operation management, decision auditing, and maintenance organization.

[0063] In some embodiments of this application, when generating evaluation results, evaluation unit 3 can perform imputation or elimination strategies based on data quality markers. Data quality markers include out-of-bounds, abrupt changes, and missing data. Evaluation unit 3 determines whether feature data can be directly used in the evaluation based on the quality markers: when feature data is missing, evaluation unit 3 can complete imputation based on the trend of feature quantities within its adjacent window; when feature data has abrupt changes or out-of-bounds, evaluation unit 3 can record it as data to be checked instead of directly using it as the evaluation basis. The processing results of imputation or elimination are recorded as data quality entries and output along with the evaluation results, enabling subsequent units to identify the reliability of the evaluation basis.

[0064] In this assessment unit 3, data quality tags are used to process and record the results during the status assessment, which achieves the following technical effects: by recording the processed results as data quality items during the assessment process, the operation status judgment process becomes traceable, enabling operation and maintenance personnel to identify data anomalies and their handling methods from the records, making the correspondence between data sources, processing strategies and assessment conclusions verifiable, ensuring consistency and interpretability of operation status analysis in long-term operation, and supporting subsequent model correction or threshold adjustment strategies.

[0065] Furthermore, after generating the evaluation results, evaluation unit 3 can output the evaluation results along with node location information. The node location information, derived from topology unit 2, indicates the location of the device corresponding to the evaluation result, as well as the structural relationship between that device and other devices in the power distribution room. This node location information allows for the differentiation between local anomalies and operational status changes that may have a wider impact range during upper-level processing.

[0066] For example, during the operation of a power distribution room, if the current change trend of a certain circuit node deviates from the historical daily curve, the evaluation unit 3 first confirms the deviation range based on the feature data provided by the acquisition unit 1, and then determines that the circuit node belongs to a certain busbar set based on the node location information provided by the topology unit 2. Subsequently, the evaluation unit 3 performs quantile modeling based on the historical daily feature data associated with the circuit node to obtain candidate warning thresholds, and verifies the candidate warning thresholds by playing back historical data. When the verification result conforms to the operating pattern, the evaluation unit 3 updates the candidate warning threshold to the current warning threshold; if the verification result does not conform to the operating pattern, the warning threshold remains unchanged, and the corresponding warning threshold retention entry is recorded.

[0067] For example, when the current characteristic exceeds the quantile modeling threshold corresponding to the historical load pattern, or when the transformer temperature, leakage current, humidity and other characteristic values ​​exceed the warning threshold range, the evaluation unit 3 can generate the corresponding abnormal evaluation result and trigger the corresponding alarm entry.

[0068] In this embodiment, the governance unit 4 is used to deduplicate, suppress, and calculate the credibility of alarms generated by the evaluation results, and to form a handling level. During this process, the governance unit 4 categorizes incoming alarm entries and establishes an index structure for the time domain and topology domain, enabling unified management of alarms of the same type, alarms from the same source, and alarms within the same topology domain. Simultaneously, the governance unit 4 maintains an alarm buffer within a time window. The buffer records alarm identifiers, timestamps, source identifiers, topology attributes, data quality markers, and historical processing status to support deduplication, suppression, and credibility calculation.

[0069] Specifically, in some embodiments of this application, the deduplication and suppression of governance unit 4 includes at least one of the following: a similar alarm merging mechanism, a backoff suppression mechanism, and a root cause merging mechanism. The similar alarm merging mechanism identifies multiple similar alarms from the same source or within the same topological domain within a time window, merges them into a single merged entry, and records the index of the original merged entry. The time window is configured by governance unit 4, and the merged entry retains the first and last timestamps, cumulative count, and representative parameters to characterize the overall performance of this type of anomaly within the window. The backoff suppression mechanism sets a cooling-off period for similar alarms within the time window. During the cooling-off period, new alarms generated from the same source or within the same topological domain do not generate new output entries but are recorded as backoff entries and referenced by the most recent valid output. The root cause merging mechanism identifies upstream key nodes based on the influence radius in the topological attributes, merges multiple downstream alarms derived from these upstream key nodes into a single root cause entry, and records the index of the merged downstream entry and the corresponding propagation association information to indicate the derivation relationship of the alarm within the topological domain.

[0070] The governance unit 4 aggregates and merges generated alarm information by utilizing alarm processing logic with time and topological correlation, and forms a controllable suppression strategy for repeatedly triggered alarms, achieving the following technical effects: First, multiple alarms caused by the same operational phenomenon are no longer presented as independent entries, but rather as a unified alarm result that can be merged. This prevents alarm outputs from being concentrated and stacked indiscriminately in a short period of time, allowing maintenance personnel to directly identify the dominant alarm source in the interface and reduce the burden of manually filtering through a large number of repetitive prompts. Second, by setting cooling and suppression conditions for alarms that are frequently triggered in a short period of time, the instantaneous disturbances, rapid fluctuations or short-term load switching in the operating data will no longer cause continuous alarm triggering. This allows the system to maintain the stability and continuity of alarm output in scenarios with large signal fluctuations or frequent switching of operating conditions, making the alarm behavior more closely reflect the actual changes in the long-term state. Third, by identifying key nodes upstream of the propagation link based on topological hierarchical relationships and merging related downstream derived alarms, the system can identify source alarms based on structural relationships, making the presentation of anomalies directional and traceable. This allows maintenance responses to no longer process downstream multi-point manifestations one by one, but to directly lock onto the source device with dominant influence, transforming the location path from multi-node search to structured directional query, thus achieving stable performance in fault location efficiency and the accuracy of handling decisions.

[0071] In some embodiments of this application, the reliability calculation of the governance unit 4 is based on weighted calculations of alarms using weighted parameters of multi-source measurement consistency, topological consistency, and sensor health, to obtain a reliability score. Multi-source measurement consistency characterizes the degree of common indication of the same event by measurement data from different channels, determined based on correlation, phase relationship, and direction of change after time alignment; topological consistency characterizes the propagation correlation between alarms under topological attributes, determined based on adjacency relationships ordered by the same busbar, adjacent loops, or influence radius; sensor health characterizes the operational stability of the acquisition source, determined based on the statistical situation of missing, abrupt, and out-of-bounds data in the data quality markers; the weighted parameters are generated by system configuration or a learning process, and the governance unit 4 writes the weighted result as a reliability score into the alarm entry; simultaneously, a mapping relationship is established between the reliability score and the handling level, which distinguishes the priority and intensity of handling actions, and the mapping relationship can be configured according to the system strategy.

[0072] In this governance unit 4, by introducing a comprehensive measurement model that integrates multi-source measurement consistency, topological consistency, and the health status of the data acquisition sources into the generated alarms, and forming a weighted credibility score, the following technical effects can be achieved: First, it enables alarm results to reflect the mutual corroboration relationship of multi-channel data in the time domain. When multiple measurement channels show a consistent trend for the same operating phenomenon, the reliability score increases. When there are differences or abnormal deviations between signals, the reliability score decreases. This makes alarms no longer based solely on single-point triggering, but forms a stable and reliable evaluation in a time-related information environment, thus making alarm performance closer to the actual operating state. Second, by placing alarms in the topology attributes for correlation analysis of propagation relationships, alarms no longer exist as isolated events, but have structural traceability. When there is a stable propagation link between an alarm and other nodes, the credibility score can reflect the position and correlation of the alarm in the topology structure, enabling the system to identify structural anomalies and making the state interpretation more directional and hierarchical, thereby improving the effectiveness of location information in operation and maintenance response. Third, the alarm results are verified by combining the health status of the data acquisition source, so that the stability of sensor operation and measurement reliability are reflected in the evaluation. When there is offset, drift or data jitter in the data acquisition source, the reliability score can be adaptively lowered to avoid false alarms caused by abnormal acquisition. This ensures that the alarm output can reflect the real working condition and forms an endogenous protection mechanism for data quality. This enables the system to maintain the consistency and reliability of the operating status assessment in long-term operation, and makes the generation process of the handling level interpretable and traceable.

[0073] Furthermore, when generating a disposal level, governance unit 4 prioritizes reading root cause entries and merged entries, and combines the credibility score and historical processing status to generate the final output. For sources or topological domains that are already in the cooling-off period, governance unit 4 does not generate new disposal level outputs, but instead updates the statistical information and latest timestamp of the existing outputs; for root cause entries, governance unit 4 includes the set of merged downstream entry associations in the output, so that subsequent processing can directly obtain the topological propagation range of the anomaly; governance unit 4 establishes a persistent record for each generated disposal level, the record content including the corresponding alarm entry index, the type of mechanism used, the credibility score, the disposal level, and the referenced historical processing status.

[0074] In some embodiments of this application, the governance unit 4 can also provide protection strategies for anomalous data. When the source of an alarm entry contains a low health status marker, the governance unit 4 assigns a lower weight to that source in the confidence calculation; when the multi-source measurement consistency of an alarm entry is insufficient, the governance unit 4 can mark it as needing review and reduce its handling level. In addition, the governance unit 4 provides an audit interface that can retrieve merged entries, backoff entries, and root cause entries by time interval, source, or topological scope, and output the associated confidence score and handling level, facilitating the tracing of the formation process of an anomaly.

[0075] For example, if a loop node generates multiple over-current alarms within a short period of time and is located on the same busbar, the governance unit 4 merges them into a single merged entry under the similar alarm merging mechanism, and establishes a backoff entry for subsequent similar alarms during the cooling period; when an abnormal temperature rise occurs in an upstream node on the same busbar and its influence radius covers the loop node, the governance unit 4 merges the over-current alarm of the loop node and the temperature rise alarm of the upstream node into a root cause entry, and then calculates a confidence score based on multi-source measurement consistency, topology consistency and sensor health, and outputs a handling level accordingly.

[0076] In this embodiment, the interlocking control unit 5 verifies the remote control command when the handling level meets the interlocking conditions, and outputs control actions according to the safety interlocking strategy, provided that personal and area safety is guaranteed. During this process, the interlocking control unit 5 correlates the handling level from the management unit 4 with the on-site safety status information to form the execution judgment conditions for the remote control command; simultaneously, it maintains the control command input queue, preprocesses each control command to be executed, and extracts the target node information and topology attributes associated with the control command, ensuring that the control judgment corresponds to the location of the equipment.

[0077] In some embodiments of this application, the interlocking control unit 5 can determine whether a control command is executable based on interlocking conditions. These interlocking conditions are defined by an interlocking strategy set, which may include protective gear wearing status, access control status, area occupancy status, and operation confirmation status, and can be configured according to the operational scenario. Specifically, the protective gear wearing status is determined by detecting the wearing status of the operator's protective equipment; the access control status is determined by reading the cabinet door magnetic sensor or access control device; the area occupancy status is identified by environmental sensing or positioning devices; and the operation confirmation status is determined by manual operation confirmation information. The interlocking control unit 5 determines whether the control command enters the execution phase based on the combination relationship between the above status variables and the handling level.

[0078] When the interlocking conditions are not met, the interlocking control unit 5 outputs a rejection message or a verification request and records an audit log. The rejection message indicates that the control command failed the interlocking condition verification, and the verification request prompts that manual confirmation or on-site verification is required. The audit log includes the timestamp of the rejection or verification, the associated node identifier, the handling level, and the relevant status variables for subsequent recording, tracing, and verification.

[0079] By introducing a multi-condition verification system related to the field status before the control command is executed and recording the interlocking execution process information, the following technical effects can be achieved in this interlocking control unit 5: First, it enables remote control actions to be correlated with on-site work behavior. When there is personnel entry, equipment operation, or work preparation, the interlocking strategy set can intercept or delay confirmation of control commands, so that the triggering of control actions is correlated with the actual work status, and the remote scheduling behavior has a consistent execution basis during operation. Second, by combining the determination of the status of protective clothing, access control and area occupancy, the control decision can reflect the dynamic changes of the environment in which the equipment is located. When there are potential risks on site or safety preparations are not completed, the interlocking conditions can form a clear interception signal, enabling remote operation to have the necessary action screening capabilities in scenarios where attention to personal risks is required, and making the control logic adaptable and continuous in different operation and maintenance scenarios. Third, by logging rejection information or review requests, the execution process of control actions has a traceable basis. Operation and maintenance personnel can identify the time, background and condition changes of interlock triggering from the records, so that remote control behavior has a clear traceability path in operation management, division of responsibilities and subsequent evaluation. This enables the system to maintain consistent control behavior with the field status during long-term continuous operation and supports the management side to dynamically adjust operation and maintenance strategies.

[0080] When the interlocking conditions are met, the interlocking control unit 5 outputs control actions according to the safety interlocking strategy. The safety interlocking strategy is used to specify the execution sequence, execution method, and execution time constraints of the control commands; the control actions include disconnection, closure, unlocking, prohibition or permission of operation, etc. After outputting the control actions, the interlocking control unit 5 establishes an execution result record, including the action execution time, action response status, and feedback signal status.

[0081] In some embodiments of this application, the interlocking control unit 5 can execute a local autonomous strategy when communication is abnormal, including: making alarm judgments based on cached warning thresholds and handling levels, limiting whitelist control actions, and generating an event log. The cached warning thresholds and handling levels are derived from the system's historical records under normal communication conditions; the whitelist control actions are pre-set by system policies and are only allowed to be executed in scenarios where the local security environment is determinable; the event log includes control judgments, action execution records, and associated state quantities generated during the autonomous period, used for reconciliation and verification after communication is restored.

[0082] After communication is restored, the interlocking control unit 5 can also perform consistency reconciliation between the event logs during the autonomous period and the cloud records. The reconciliation process includes comparing the event logs during the autonomous period with the operation records in the cloud one by one, and determining whether there are any execution deviations. After the consistency reconciliation is completed, the interlocking control unit 5 performs a playback verification, matching the control records during the autonomous period with the evaluation data, governance data, and topology attributes, thereby restoring the system to a complete and consistent state after experiencing communication anomalies.

[0083] In the case of communication failure, the interlocking control unit 5 executes a local autonomous strategy. By using cached information during autonomous operation to maintain the executability of alarm judgment and control actions, and performing consistency reconciliation and playback verification after communication is restored, the following technical effects can be achieved: First, it enables the system to maintain basic operational capabilities under conditions of communication link interruption, delay, or fluctuation. The autonomous strategy uses the cached warning threshold and handling level as the basis for judgment, so that alarm judgment does not depend on the real-time uplink, the feedback of the operating status is not interrupted, and the equipment can still maintain continuous operation even without cloud support. Second, by limiting whitelist control actions, the control behavior during the autonomous period is kept within a verified and safe execution range, avoiding control commands that do not meet the operating conditions due to control link mismatch caused by communication anomalies. This makes the execution under the autonomous strategy predictable and safe and controllable, and enables the operating organization to have edge stability in the remote disconnection state. Third, after communication is restored, consistency reconciliation and playback verification of the event logs and cloud records formed during the autonomous period are performed to ensure that the operational behavior, alarm changes and control execution during the autonomous period form an auditable complete link. This enables the system to confirm the consistency between autonomous behavior and normal operation behavior, and to continuously express the trend of state changes in the operation records. This provides reliable support for the consistency, traceability and executive execution of the control process during long-term system operation.

[0084] For example, when the power distribution room is under maintenance and there are on-site workers, the protective equipment status and the area occupancy status are indicated by corresponding status variables. When the interlocking control unit 5 receives a remote closing command, it first determines whether the interlocking conditions are met. If the protective equipment status or the area occupancy status is not met, the interlocking control unit 5 outputs a rejection message and records an audit log. When all interlocking conditions are met, the interlocking control unit 5 executes the closing action according to the control strategy and records the execution result.

[0085] In this embodiment, the operation and maintenance unit 6 is used to generate preventive test plans and maintenance task arrangements based on topology attributes and handling levels, and update the early warning thresholds, handling levels, and safety interlocking strategies based on the test results. During this process, the operation and maintenance unit 6 receives the handling level output by the governance unit 4 and the topology attributes output by the topology unit 2, and, combined with the structural positional relationship of the equipment in the local topology model, determines the target objects that should be prioritized for inspection or maintenance; simultaneously, it maintains preventive test records and maintenance task status records, and updates the parameters of the evaluation unit 3 and the interlocking control unit 5 after the tasks are completed, to ensure that the operational status assessment and safety control strategies remain consistent and effective under long-term operating conditions.

[0086] In some embodiments of this application, when generating a preventative test plan, the operation and maintenance unit 6 can screen and sort target objects according to the influence radius ranking and the disposal level. The influence radius ranking is used to characterize the topological propagation hierarchy between target objects; the disposal level is used to distinguish the risk level of target objects in their current operating state. The operation and maintenance unit 6 selects target objects at the top of the ranking based on the influence radius ranking and the disposal level, enabling preventative tests to preferentially cover nodes that may have a transmissive impact or are located in high-risk areas.

[0087] After selecting the target objects, the operation and maintenance unit 6 generates a test execution path for the preventive test plan based on the interdependencies between the target objects. This path characterizes the inspection sequence, adjustment conditions, and priority of the target objects. Specifically, the inspection sequence is arranged from upstream to downstream nodes according to the topology propagation direction and treatment level; the adjustment conditions indicate whether the subsequent inspection sequence needs to be changed if a target object exhibits a specific operational state during the test; the priority characterizes the order in which multiple target objects are tested at the same level; and the preventive test plan includes at least one of the following: target object identifier, inspection items, execution sequence, and completion status, used for scheduling, tracking, and recording during execution.

[0088] In this operation and maintenance unit 6, preventive test plans with execution order are generated based on topology attributes and treatment levels. Test objects are selected and paths are constructed based on influence radius ranking, achieving the following technical effects: First, it enables the generation of test plans to no longer rely on experience-based judgment or static periodic arrangements, but to be dynamically organized based on the structural relationships and impact range of the current operating status. This makes the selection of test objects interpretable and allows operation and maintenance activities to be carried out around nodes that may have a chain reaction effect, thereby forming a continuous relationship between test arrangements and operating status. Second, by constructing an experimental execution path that includes the testing sequence and adjustment conditions, the experimental behavior is no longer a discrete action executed independently, but has a logical connection between nodes based on the dependency relationship. When the state of a target object changes, the execution path can be adjusted according to the set conditions, so that the experimental process can maintain the integrity and consistency of the execution chain in different operating scenarios. Third, by including information such as target object identification, inspection items, execution sequence, and completion status in the test plan, the execution process of preventive tests is recordable and traceable. Maintenance personnel can clearly identify the test progress from the plan and execution records, and the maintenance behavior has a clear record link at the management and audit level. This ensures a consistent correspondence between equipment status changes and test execution records in long-term operation and maintenance, and makes the organization and results of operation and maintenance stable and continuous.

[0089] In the maintenance task scheduling section, the operation and maintenance unit 6 can generate a maintenance task list based on the inspection items and results of each target object in the test execution path. This list indicates the verification, adjustment, or replacement operations that need to be performed. The maintenance task scheduling records the allocation, execution, and completion status of each task and links it to the equipment lifecycle management information, enabling the operation and maintenance process to form a continuously updated closed loop of equipment health information.

[0090] Simultaneously, upon receiving the test results, the maintenance unit 6 updates the warning threshold, handling level, and safety interlock strategy. Test results may include equipment insulation test results, partial discharge detection results, temperature rise detection results, or operational status verification results. The maintenance unit 6 compares the test results with the operational history maintained by the evaluation unit 3 and adjusts the warning threshold appropriately based on the test results, ensuring that the warning threshold reflects the trend of equipment status changes. For target objects with significant operational risks, the maintenance unit 6 raises the corresponding handling level; for target objects whose test results indicate a stable state, the maintenance unit 6 can maintain or lower their handling level.

[0091] When updating the safety interlocking strategy, the maintenance unit 6 adapts and adjusts the protective equipment status, access control status, area occupancy status, and operation confirmation status in the interlocking strategy set based on the test results and maintenance task execution status, so that the safety interlocking strategy can match the equipment operating status during the test and maintenance processes. The maintenance unit 6 establishes a persistent record of each test plan generation, maintenance task arrangement, and strategy update, and can retrieve these records based on time intervals, target objects, or topology levels.

[0092] For example, when a busbar node in the substation is identified as a high-level issue in the management unit 4, the operation and maintenance unit 6 can select downstream loop nodes from the influence radius ranking as target objects, generate a preventative test plan, which may include test items such as insulation, current temperature rise, and partial discharge, or test items such as winding DC resistance, insulation resistance, absorption ratio, or polarization index, and determine the test execution order based on the interrelationship between target objects. After receiving the results of each test item, the operation and maintenance unit 6 updates the warning threshold and issue level of the busbar node and its corresponding downstream loop nodes, realizing continuous synchronization between the operating status assessment parameters and the actual equipment status.

[0093] In summary, this system, through the collaborative efforts of acquisition unit 1, topology unit 2, evaluation unit 3, governance unit 4, interlocking control unit 5, and operation and maintenance unit 6, continuously acquires the operating data of power distribution equipment, constructs structural correlations, and comprehensively analyzes operating and historical characteristics. This enables timely identification of abnormal states, effective organization of alarm information, and safe execution of control commands. Furthermore, during maintenance, test results are written back for updating operating strategies, ensuring consistency in state judgment, reliability in control execution, and coherence in maintenance decisions during long-term operation of power distribution equipment.

[0094] Example 2: like Figure 2 As shown in the figure, this embodiment provides a method for monitoring and fault early warning of intelligent power distribution equipment, which includes the following steps: Step S1: Obtain the electrical parameters, environmental parameters, and equipment operating status of the power distribution equipment, and generate time-synchronized characteristic data; Step S2: Generate a local topology model of the power distribution room based on the relationship between the cabinets, circuits and busbars of the power distribution equipment, and output the topology attributes corresponding to the feature data; Step S3: Generate an evaluation result of the operating status based on feature data and topology attributes, and update the warning threshold based on the preset warning threshold according to the operating history and load pattern; Step S4: Generate alarms based on the evaluation results, and perform deduplication, suppression, and confidence calculation on the alarms to obtain the handling level; Step S5: When the handling level meets the interlocking conditions, verify the remote control command, and output control actions according to the safety interlocking strategy under the premise of ensuring personal and area safety; Step S6: Generate a preventative test plan and maintenance task schedule based on topology attributes and response levels, and update the warning threshold, response level, and safety interlock strategy based on the test results.

[0095] By adopting the above technical solution, a coherent response relationship is formed between status identification, alarm handling, control execution and maintenance adjustment within the system, realizing a closed expression of the operating status from monitoring to execution and then to backtracking. This enables the operating information to maintain stable indication characteristics under conditions of load change, environmental change and equipment aging, and ensures that control actions are consistent with the field status.

[0096] Specifically, the state representation composed of feature data and structural attributes provides relevant input for subsequent judgments; the introduction of operating history and load patterns enables early warning judgments to self-adjust over time; alarms undergo correlation aggregation processing before entering the handling chain, giving the presented results a centralized direction; control commands undergo on-site state verification before execution, ensuring the actions themselves have verifiable premises; maintenance activities are organized based on state indications and structural relationships, allowing the maintenance rhythm to be adjusted according to changes in operating status; through the operation of the above-mentioned correlation chains, the power distribution system presents a continuous, reviewable, and deducible state change path during operation.

[0097] Hereinafter, steps S1 to S6 will be specifically described according to embodiments of this application.

[0098] In step S1, it is necessary to acquire the electrical parameters, environmental parameters, and equipment operating status parameters of the power distribution equipment and form time-synchronized feature data. The electrical parameters may include voltage, current, active power, reactive power, power factor, three-phase imbalance, and phase-to-phase current difference, etc.; environmental parameters may include temperature, humidity, water immersion, smoke, and cabinet access control status, etc.; equipment operating status parameters may include circuit breaker opening and closing positions, relay protection action signals, secondary circuit alarm signals, and transformer temperature rise protection trigger signals, etc. The above data can come from different measuring devices and monitoring points. The acquisition device timestamps the data from different sources and aligns them according to a unified time reference, enabling data from different sampling periods to form a corresponding relationship within the same time window. In the aligned data, the original measured values ​​are retained, and derived quantities are constructed based on feature extraction logic, such as phase-to-phase difference, phase sequence offset features, and temperature change rate features, to reflect the changing trend of the equipment operating status. Meanwhile, missing, abrupt, and out-of-bounds data are marked to identify the source of data quality issues in subsequent steps and to differentiate the data accordingly.

[0099] In step S2, a local topology model of the power distribution room is generated based on the relationships between the cabinets, circuits, and busbars of the power distribution equipment, and the topology attributes corresponding to the feature data are output. First, a node set is established based on the equipment installation information and outgoing circuit configuration of the power distribution room. The node set includes cabinet nodes representing cabinet units, circuit nodes representing outgoing branches, and busbar nodes representing electrical convergence points. Then, based on the structural relationship between circuit nodes and cabinet nodes, and the electrical connection relationship between circuit nodes and busbar nodes, the connection relationships between nodes are constructed, forming a topology representation. The topology model can be structured using a node table and a connection table, allowing for clear positioning of the upstream and downstream directions of any node. For the feature data generated in step S1, a mapping is established between the collection point identifier and the node identifier, ensuring that each feature data has corresponding topology attribute information, including at least the node's hierarchical position in the topology, its busbar identifier, and upstream / downstream propagation direction identification information. Based on this, an influence radius can be generated based on the topology structure to represent the hierarchical distance relationship between any node and the target node. The influence radius sorting is used to represent the order in which the influence of a node may propagate in the topology, providing a structural basis for subsequent anomaly source identification and handling path planning.

[0100] In step S3, an evaluation result of the operating status is generated based on feature data and topology attributes. The warning threshold is then updated based on the preset warning threshold, according to the operating history and load pattern. First, the feature data output in step S1 is associated with the topology attributes output in step S2, ensuring each feature data point has a unique location in the structural space. The operating status evaluation can be based on interval determination using the warning threshold, which can be set separately for different equipment objects. The operating history includes time-series records of feature data, and the load pattern can be constructed based on daily and weekly patterns to model the distribution of feature quantities in different time periods. During the threshold update process, quantile modeling is performed on the feature quantities to obtain candidate warning thresholds. The candidate warning thresholds are then reviewed against the operating history. The evaluation result is compared with the recorded operating events to determine if the candidate warning threshold is consistent with historical operating patterns. If the review result meets the consistency requirement, the warning threshold is updated and the changed item is recorded. If the review result does not meet the consistency requirement, the current warning threshold is maintained and the maintained item is recorded. When data is missing or fluctuates abnormally, imputation or elimination is performed based on data quality markers, and the processing results are output as data quality items along with the evaluation results.

[0101] In step S4, alarms are generated based on the evaluation results, and deduplication, suppression, and confidence calculation are performed on the alarms to obtain the handling level. First, the abnormal state judgment results generated in step S3 are aggregated according to device object, time window, and topology location to form alarm entries, including evaluation status, timestamp, node identifier, and data quality marker. When performing deduplication and suppression on alarms, alarms with the same source or topology scope within the time window are identified, merged into a single entry, and the number of merges and the first and last timestamps are recorded. In the backoff suppression process, a cooling-off period is set for alarms of the same type; during the cooling-off period, new alarms are not repeatedly output, but are marked as backoff entries. In the root cause merging process, nodes located upstream in the topology with a small influence radius are identified as root cause nodes, and their downstream derived alarms are merged into root cause alarm entries. The confidence calculation is weighted based on multi-source measurement consistency, topology consistency, and the health of the acquisition source to obtain a confidence score. The handling level is mapped based on the confidence score and the degree of abnormality for subsequent control strategy applications.

[0102] In step S5, the remote control command is verified when the handling level meets the interlocking conditions, and control actions are output according to the safety interlocking strategy, provided that personal and area safety is guaranteed. The interlocking conditions consist of protective gear status, access control status, area occupancy status, and operation confirmation status, representing the safety preconditions before executing the control command. First, the control command is parsed to determine its corresponding target object and scope of action; then, the handling level of the target object is compared with the interlocking conditions to determine whether to enter the control execution phase. When any interlocking condition is not met, a rejection record or review request is generated; when the interlocking condition is met, control actions are output according to the safety interlocking strategy, and the control execution result is recorded. In case of communication failure, the local autonomy strategy is executed, the execution scope is determined based on the locally cached warning threshold and handling level, and consistency reconciliation and playback verification are performed on the records during the autonomy period after communication is restored.

[0103] In step S6, a preventative test plan and maintenance task arrangement are generated based on topology attributes and response levels. The warning threshold, response level, and safety interlock strategy are then updated based on the test results. First, based on target objects with higher response levels, and sorted by influence radius, objects with topological propagation relationships are selected to form a test target set. Then, test execution paths are generated based on the relationships and execution dependencies between objects. These paths include the order of test objects, execution conditions, and priorities. The test plan may include tests such as insulation, temperature rise, partial discharge, or electrical withstand voltage. After the test is completed, the warning threshold is adjusted based on the test results, the response level is maintained or modified, and the operating conditions involved in the safety interlock strategy are updated synchronously to ensure consistency between system judgments and equipment status during long-term operation.

[0104] In summary, this method obtains the operating data of power distribution equipment and establishes a structural association with the topological relationship. Combined with operating status assessment and threshold updates, it organizes, filters, and determines the credibility of abnormal alarms. Under the premise of meeting safety conditions, it executes control commands and writes back the test results for updating operating parameters and strategies. This achieves a closed-loop processing chain from monitoring, judgment, handling to maintenance, enabling power distribution equipment to maintain the continuity of status identification and the coordination of control execution during long-term operation.

[0105] It should be noted that, although the embodiments in this application are based on... Figure 1 Steps S1 to S6 are described sequentially, but this does not mean that steps S1 to S6 must be performed in a strict order. The reason this embodiment follows this order is... Figure 1The order in which steps S1 to S6 are described is provided to facilitate understanding of the technical solutions of the embodiments of this application by those skilled in the art. In other words, in the embodiments of this application, the order of steps S1 to S6 can be appropriately adjusted according to actual needs.

[0106] The above-disclosed embodiments are merely preferred embodiments of the present invention, but the present invention is not limited thereto. Any non-creative variations that can be conceived by those skilled in the art, as well as any improvements and modifications made without departing from the principles of the present invention, should fall within the protection scope of the present invention.

Claims

1. A smart power distribution equipment monitoring and fault early warning system, characterized in that, include: The data acquisition unit is used to acquire electrical parameters, environmental parameters, and equipment operating status data of the power distribution equipment, and to generate time-synchronized characteristic data. A topology unit is used to generate a local topology model of the power distribution room based on the relationship between the cabinets, circuits and busbars of the power distribution equipment, and output the topology attributes corresponding to the feature data. An evaluation unit is used to generate an evaluation result of the operating status based on the feature data and the topology attributes, and to update the early warning threshold based on the operating history and load pattern, on the basis of a preset early warning threshold. The governance unit is used to deduplicate, suppress, and calculate the credibility of alarms generated from the evaluation results, and to form a handling level. The interlocking control unit is used to verify the remote control command when the handling level meets the interlocking conditions, and to output control actions according to the safety interlocking strategy under the premise of ensuring personal and area safety; The operation and maintenance unit is used to generate preventive test plans and maintenance task arrangements based on the topology attributes and the handling level, and to update the early warning threshold, the handling level and the security interlocking strategy based on the test results.

2. The intelligent power distribution equipment monitoring and fault early warning system as described in claim 1, characterized in that, When generating the local topology model, the topology unit represents the local topology model in a structured way with nodes and connection relationships and generates topology attributes. The topology attributes include the influence radius of each node relative to the target object. The influence radius is used to characterize the propagation level of the anomaly from any node to the target object in the local topology model. The topology unit generates an influence radius sorting based on the influence radius to indicate the propagation order and propagation range of the nodes.

3. The intelligent power distribution equipment monitoring and fault early warning system as described in claim 1, characterized in that, The step of updating the warning threshold based on operating history and load pattern includes: performing quantile modeling on the feature data based on the daily and weekly patterns in the load pattern to generate candidate warning thresholds; verifying the candidate warning thresholds by playing back historical data; updating the warning threshold with the candidate warning thresholds and generating corresponding warning threshold change entries when the verification passes; and keeping the warning threshold unchanged and generating warning threshold retention entries when the verification fails.

4. The intelligent power distribution equipment monitoring and fault early warning system as described in claim 2, characterized in that, The process of deduplicating and suppressing alarms generated by the evaluation results includes: using at least one of a similar alarm merging mechanism, a backoff suppression mechanism, and a root cause merging mechanism for deduplication and suppression. The similar alarm merging mechanism merges alarms from the same source or the same topological domain within a time window. The backoff suppression mechanism sets a cooling-off period for similar alarms within the time window. The root cause merging mechanism merges alarms derived from upstream key nodes within the influence radius into a single root cause entry.

5. The intelligent power distribution equipment monitoring and fault early warning system as described in claim 1, characterized in that, The credibility calculation includes: calculating the credibility score of the alarm based on the weight parameters of multi-source measurement consistency, topology consistency and sensor health. The multi-source measurement consistency is used to characterize the correlation of multi-channel measurement data in the time domain. The topology consistency is used to characterize the propagation correlation of the alarm in the topology attribute. The sensor health is used to characterize the working stability of the acquisition source. The weight parameters are generated by system configuration or learning process. The credibility score is associated with the handling level.

6. The intelligent power distribution equipment monitoring and fault early warning system as described in claim 1, characterized in that, The interlocking conditions are defined by an interlocking strategy set, which includes protective clothing status, access control status, area occupancy status, and operation confirmation status, and can be configured according to the operating scenario. When the interlocking conditions are not met, the interlocking control unit outputs a rejection message or a review request and records an audit log.

7. The intelligent power distribution equipment monitoring and fault early warning system as described in claim 1, characterized in that, When communication is abnormal, the interlocking control unit executes a local autonomous strategy, which includes: making an alarm determination based on the cached warning threshold and the handling level, limiting whitelist control actions and generating an event log; after communication is restored, reconciling the event log during the autonomous period with the cloud records and performing a replay verification.

8. The intelligent power distribution equipment monitoring and fault early warning system as described in claim 2, characterized in that, The step of generating a preventive test plan based on the topology attributes and the treatment level includes: selecting target objects that are in the order of the influence radius sorting and the treatment level, and generating test execution paths for the preventive test plan based on the interdependencies between the target objects. The test execution paths identify the inspection order, adjustment conditions, and priority of the target objects. The preventive test plan includes at least one of the following: target object identifier, inspection items, execution order, and completion status.

9. The intelligent power distribution equipment monitoring and fault early warning system as described in claim 1, characterized in that, The acquisition unit performs time synchronization and data quality marking. The data quality marking is used to indicate the anomaly type, which includes at least one of out-of-bounds, mutation, and packet loss. When generating the evaluation result, the evaluation unit performs an interpolation or elimination strategy based on the data quality marking and records the processing result as a data quality entry.

10. A method for monitoring and fault early warning of intelligent power distribution equipment, characterized in that, Includes the following steps: Acquire electrical parameters, environmental parameters, and equipment operating status parameters of power distribution equipment, and generate time-synchronized characteristic data; A local topology model of the power distribution room is generated based on the relationship between the cabinets, circuits and busbars of the power distribution equipment, and the topology attributes corresponding to the feature data are output. An evaluation result of the operating status is generated based on the feature data and the topology attributes, and the warning threshold is updated according to the operating history and load pattern based on the preset warning threshold. An alarm is generated based on the evaluation results, and the alarm is deduplicated, suppressed, and its credibility is calculated to obtain the handling level. When the handling level meets the interlocking conditions, the remote control command is verified, and control actions are output according to the safety interlocking strategy under the premise of ensuring personal and area safety. Based on the topology attributes and the response level, a preventative test plan and maintenance task schedule are generated, and the warning threshold, the response level, and the safety interlock strategy are updated based on the test results.

Images