A power distribution network equipment health degree evaluation and intelligent operation and maintenance method based on negative area integral model

Abstract

The application discloses a kind of distribution network equipment health degree evaluation and intelligent operation and maintenance method based on negative area integral model.The method is by building the dynamic negative area integral calculation model of fusion equipment outage, operating abnormity and multi-source data such as customer complaint, realize the accurate quantitative evaluation of the health state of line, distribution transformer and other equipment, and generate equipment panoramic health portrait for visual early warning.System is based on the evaluation result automatically triggers or intelligently recommends the operation and maintenance strategy such as differentiated patrol, maintenance, and pushes to the scene execution through mobile application.At the same time, the integral model is continuously self-optimized by using operation and maintenance closed-loop feedback data.The application effectively solves the problems of data dispersion, evaluation lag and decision-making relying on experience in traditional operation and maintenance, improves the initiative, accuracy and intelligent level of distribution network operation and maintenance, and guarantees power supply reliability and service quality.

Description

Technical Field

[0001] This invention relates to the fields of smart grid and digital operation and maintenance technology, and in particular to a method for health assessment and intelligent operation and maintenance of distribution network equipment based on a negative integral model. Background Technology

[0002] The distribution network is the final link connecting the main power grid and a large number of power users, and the reliability of its equipment operation directly affects the power supply quality and user satisfaction. With the continuous expansion of the power grid and the increasing demands of users, traditional operation and maintenance management models are no longer able to meet the requirements of refined and intelligent management.

[0003] Currently, the operation and maintenance management of power distribution networks heavily relies on multiple independently constructed business information systems, such as electricity consumption information collection systems, distribution automation systems, and marketing management systems. These systems differ in their construction period, technical standards, and data specifications, resulting in the fragmented storage and inconsistent formats of critical operational data, power outage events, and customer work orders. This data fragmentation leads to fragmented equipment status information, making it impossible to form a unified view of equipment health. For example, the impact range of a distribution network fault, related equipment, and user complaints cannot be analyzed and presented in real time, making it difficult for operation and maintenance personnel to quickly and accurately grasp the overall situation.

[0004] At the business functionality level, existing systems focus on recording basic information and post-event statistics, lacking in-depth assessment of equipment health status and proactive early warning capabilities. For example, in power outage management, existing functions are mostly limited to event recording and simple classification, unable to perform multi-dimensional root cause analysis on equipment with frequent power outages, nor can they automatically verify and assess the quality of power outage events based on real-time measurement data. Furthermore, there is a lack of dynamic linkage and consistency verification mechanisms between equipment ledgers, topology relationships, and real-time data, making it difficult to promptly detect and address data distortion caused by equipment anomalies, directly impacting the accuracy of fault diagnosis and emergency repair command.

[0005] Furthermore, the existing technological system has significant shortcomings in supporting cross-departmental and cross-professional collaborative operations. Taking business expansion application as an example, the process involves multiple departments such as marketing, equipment, and planning, but key node information is not uniformly transferred and transparently monitored online, making it difficult to track project progress, resulting in low efficiency in inter-departmental collaboration and affecting the customer's power connection time. In terms of management decision support, the existing system provides relatively simple analytical functions and insufficient visualization, failing to integrate multi-source data to build a comprehensive profile of equipment, customers, or power supply stations. Management lacks intuitive and efficient data insight tools to support precise resource allocation and strategic planning.

[0006] In summary, existing technical solutions have limitations in terms of data integration, business intelligence, collaborative efficiency, and decision support depth, which hinder further improvements in the quality and efficiency of power distribution network operation and maintenance and service levels. Therefore, there is an urgent need for a systematic solution that can achieve data connectivity, intelligent assessment, collaborative management and control, and decision optimization. Summary of the Invention

[0007] To address the aforementioned technical problems, this invention provides a method for health assessment and intelligent operation and maintenance of distribution network equipment based on a negative integral model.

[0008] The specific details of the invention are as follows:

[0009] A method for health assessment and intelligent operation and maintenance of distribution network equipment based on a negative integral model includes:

[0010] S1. Model Construction: For lines and distribution transformers in the distribution network, a negative integral calculation model is constructed and maintained respectively. The model defines multiple integral factors, the weight of each factor, the rules for integral accumulation and operation and maintenance score elimination, and the integral threshold corresponding to different health levels.

[0011] S2. Data-driven point calculation: Real-time access to multi-source data from electricity consumption information collection, power distribution automation, dispatch automation and marketing service systems, and dynamic calculation and updating of negative point values ​​for each device based on the negative point calculation model;

[0012] S3. Health Assessment and Visualization: Based on the real-time negative score value of the equipment, its health level is determined, and a health profile of the equipment that integrates score details, historical events and ledger information is generated and visualized and alerted on a power grid map.

[0013] S4. Intelligent Operation and Maintenance Strategy Generation: Based on the health level and profile information of the equipment, automatically trigger or assist in generating differentiated operation and maintenance work orders, maintenance plan suggestions or resource allocation schemes.

[0014] Preferably, the integral factors in the negative integral calculation model include at least: equipment downtime frequency and duration factor, equipment electrical measurement anomaly factor, and related customer complaint work order factor; wherein, the weight of the customer complaint work order factor is dynamically adjusted according to the authority of the complaint source channel and the severity of the complaint issue.

[0015] Preferably, the method further includes an adaptive optimization step for the integral factor: performing a correlation analysis based on the trend of equipment integral changes within a historical period and subsequent failure events, and dynamically optimizing the weight configuration of the integral factor using a machine learning algorithm.

[0016] Preferably, in step S3, generating a device health profile specifically includes: establishing a point contribution tracing mechanism to thoroughly examine the details of the main events that lead to the current negative points; at the same time, integrating the results of device topology consistency verification into the profile to directly associate data anomalies with device health assessment.

[0017] Preferably, in step S4, generating differentiated operation and maintenance strategies specifically includes:

[0018] For equipment that reaches the warning threshold, a list of negative issues for that equipment will be automatically added to the planned inspection work order to guide on-site personnel to focus on inspections.

[0019] In proactive repair or fault reporting scenarios, if the work order involves equipment with high negative scores, the historical anomaly records and handling suggestions for that equipment will be pushed to the repair personnel on the mobile application.

[0020] The negative points of equipment will be used as an important basis for decision-making in the establishment and reserve of distribution network projects, and priority will be given to the transformation of high-frequency negative equipment.

[0021] Preferably, the method further includes a closed-loop feedback mechanism: tracking and recording the processing results of maintenance work orders triggered by negative points, and deducting equipment points according to the point reduction rules if the problem is confirmed to be resolved; at the same time, the processing results are fed back to the point calculation model as sample data to optimize the accuracy of points and point reduction rules.

[0022] A distribution network equipment health assessment and intelligent operation and maintenance system for implementing the above method includes:

[0023] Model Management Center: Provides a graphical interface for configuring and maintaining negative score calculation models for different device types, including defining factors, weights, rules, and thresholds;

[0024] Real-time integration calculation engine: integrated with the power grid resource business platform, measurement center and marketing system, it acquires data in real time and drives the integration model to perform calculations, outputting dynamic integration of equipment;

[0025] Equipment Health Profile Service: Receives points data, generates and manages panoramic health profiles of equipment, and provides multi-dimensional query, traceability analysis and visualization services;

[0026] Intelligent operation and maintenance strategy service: Based on the health information output by the profile service and combined with the pre-set strategy library, it automatically generates or recommends specific operation and maintenance action plans and pushes them to the power supply service command system work order process or mobile application.

[0027] Preferably, the system further includes a model self-optimization module, which automatically adjusts the parameters of the negative integral calculation model periodically or triggeredly based on the data sequence of "integral event handling results" accumulated by the operation and maintenance closed-loop feedback mechanism.

[0028] Preferably, the model self-optimization module performs the following steps to optimize the negative integral calculation model:

[0029] a) Multi-feature contribution analysis: Extract and combine features of various events that lead to an increase in equipment score in historical data, and quantify the contribution fluctuation of each score factor to the final score result under different operating scenarios.

[0030] b) Gradual weight adjustment: Based on the contribution fluctuation obtained in step a), and combined with the effective handling rate of recent maintenance work orders, the weights of each integral factor in the model are adjusted non-linearly and gradually to make the weight allocation more in line with the actual risk level.

[0031] c) Timing pattern recognition and rule fine-tuning: Identify the periodic abnormal change patterns of device integrals under specific seasons, weather or load modes, and make adaptive fine-tuning of the time decay coefficient or event aggregation conditions in the integral accumulation rules accordingly.

[0032] d) Grouping and Differentiation Feedback: The equipment is grouped according to type, region, and years of operation. The differences in the correlation between integral changes and failure occurrence among different groups are analyzed, and the differences are fed back to the weight correction process in step b) to achieve group-differentiated adaptation of model parameters.

[0033] e) Iterative evaluation of strategy value: Establish an evaluation mechanism to retrospectively evaluate the actual effect of operation and maintenance strategies triggered by the optimized model, and use the evaluation results as the target guide for the next round of model parameter optimization.

[0034] Preferably, the multi-feature contribution analysis in step a) adopts a dynamic feature cross method, which not only analyzes the independent contribution of a single integral factor, but also focuses on analyzing the synergistic amplification or suppression effect on the integral when two or more factors appear simultaneously under specific spatiotemporal conditions, and quantifies this effect into a dynamic synergistic correction coefficient for more refined integral calculation.

[0035] The beneficial effects of this invention are as follows: by constructing a dynamic negative integral model based on multi-source data fusion, it achieves accurate quantitative assessment and visualization of the health status of distribution network equipment, effectively solving the problems of inaccurate equipment status perception and delayed early warning caused by data dispersion; by intelligently linking the assessment results with the operation and maintenance process, it automatically generates differentiated operation and maintenance strategies and pushes them to the field, significantly improving the initiative and accuracy of operation and maintenance operations, and reducing the burden of manual judgment and operation and maintenance costs; at the same time, relying on the closed-loop feedback and model self-optimization mechanism, the assessment model can continuously adapt to changes in the power grid and continuously optimize the accuracy of early warning, thereby enhancing the overall reliability of power supply in the distribution network, customer service satisfaction, and the scientific nature of management decisions. Attached Figure Description

[0036] Appendix Figure 1 System overall architecture diagram;

[0037] Appendix Figure 2 Flowchart of the algorithm for real-time calculation of negative scores and health assessment;

[0038] Appendix Figure 3 : Schematic diagram of the negative integral model library structure;

[0039] Appendix Figure 4 Figure 1 shows a diagram of a latent defect early warning system based on long-term trend identification. Figure 2 shows the trend of multi-factor scores and total score, and Figure 3 shows a diagram of a mobile early warning work order.

[0040] Appendix Figure 5 : Schematic diagram of the working logic of the model self-optimization module. Detailed Implementation

[0041] The following embodiments illustrate the present invention in detail. In the description of these embodiments, specific details such as particular system structures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods are omitted so as not to obscure the description of this application with unnecessary detail.

[0042] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0043] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0044] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0045] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0046] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0047] Example 1

[0048] Reference Appendix Figure 1 With appendix Figure 2 This embodiment details the overall architecture, core algorithms, and complete workflow of a distribution network equipment health assessment and intelligent operation and maintenance system based on a negative integral model. As an enhancement component of the power supply service command system, this system aims to achieve quantitative assessment of distribution network equipment status and intelligent operation and maintenance decision-making.

[0049] I. System Architecture and Data Flow

[0050] The system adopts a layered microservice architecture, which mainly includes the following layers and components:

[0051] 1. Unified Data Access and Governance Layer: This layer serves as the foundation of the system's data, establishing real-time data channels with eight core external source systems through an enterprise-level service bus and standard API interfaces. Specifically, it includes:

[0052] The system uses data from power outages and power surges of public and private transformers, as well as measurement curves for voltage, current, and active power, to access the 2.0 system.

[0053] Access switch tripping information, protection action signals, line flow diagrams, and real-time operating data from the power distribution automation system.

[0054] The Marketing 2.0 system integrates business expansion work orders, non-emergency repair work orders, customer power outage procedures due to unpaid bills, customer basic information, and 95598 complaint work orders.

[0055] Access the main grid switch trip signal (SOE event), main grid measurement curves, and protection information from the dispatch automation system.

[0056] Access the power distribution network engineering equipment modification information and completion acceptance status from the PMS3.0 system.

[0057] Access the operation data and events of transmission lines and main transformers from the enterprise-level measurement center.

[0058] Access real-time meteorological data and early warning information from enterprise-level meteorological centers.

[0059] Obtain unified equipment ledgers, topology relationships, and "one map of the power grid" services from the power grid resource business platform.

[0060] This layer has built-in data cleaning, formatting, and standardization modules to ensure that heterogeneous data from multiple sources has consistent specifications and quality before entering the core engine.

[0061] 2. Negative Integral Model Library and Computation Engine Layer: This is the core of the system's intelligence.

[0062] Model Library: Stores predefined negative integral calculation models for different equipment types (such as 10kV lines, pole-mounted transformers, ring main units, etc.). Each model is a configurable rule set that explicitly defines:

[0063] Integral Factors: Define which events or states will affect the points. Core factors include, but are not limited to: outage factors (planned power outages, power failures), electrical anomaly factors (voltage / current exceeding limits, imbalance, load factor), customer impact factors (complaint work order level and number), equipment anomaly factors (inconsistent records, topology errors), and environmental factors (number of times severe weather affects the points).

[0064] Factor weights: The proportion of each factor's contribution to the total score, which can be configured differently based on the importance of the equipment and the characteristics of the region.

[0065] Scoring mapping rules: Rules that convert raw events (such as a power outage lasting 2 hours) into specific scores (such as 5 points).

[0066] Integral aggregation and decay algorithm: specifies how to accumulate the scores of multiple events, and how the scores of old events decay over time.

[0067] Synergy effect rule: Defines whether additional points are awarded when multiple related factors (such as "power outage" and "a large number of complaints during the same period") occur simultaneously.

[0068] Health threshold: Divide the score ranges corresponding to different health levels such as "healthy", "sub-healthy", "concerned", and "serious".

[0069] Real-time computing engine: A high-performance stream processing service. It continuously monitors the event stream of the data access layer, and once an event related to any integration factor is detected, it immediately triggers a recalculation of the integration on the affected devices. The calculation process strictly follows the rules defined in the model library to ensure the timeliness and accuracy of the integration.

[0070] 3. Intelligent Application Service Layer: Provides various business applications based on the points results.

[0071] Device Health Profile Service: Generates a dynamic, panoramic health profile for each device. This profile not only includes the current total score and health level, but also displays the composition of the score (the contribution ratio of each factor) in a visual way (such as pie charts and trend lines), and supports step-by-step viewing of details of every historical event that caused the score change. This service is deeply integrated with the "Power Grid Map," supporting the intuitive display of the health distribution of devices across the entire region on a geographic map using color rendering (such as green for healthy and red for severe).

[0072] Intelligent O&M Strategy Engine: Embedded with a policy rule base, it maps device health status to specific O&M suggestions or instructions. For example:

[0073] Rule 1: If the device health status changes to "Attention" and the main factor is "Frequent Power Outages", a "Special Inspection Work Order" will be automatically generated and pushed to the mobile APP.

[0074] Rule 2: If the equipment health status is "critical" and it is associated with an "important user", the "proactive maintenance command" process will be triggered immediately, and the account manager will be notified.

[0075] Rule 3: If multiple devices on a certain line simultaneously enter a "sub-healthy" state, a "line comprehensive maintenance" project reserve suggestion will be generated.

[0076] Model Self-Optimization Service: This is a background learning module. It continuously collects complete closed-loop data from "equipment score changes -> triggering maintenance actions -> on-site handling results". By analyzing a large amount of closed-loop data, this service evaluates the effectiveness of the current score model (such as early warning accuracy) and uses algorithms (such as gradient descent-based weight tuning and correlation analysis-based rule discovery) to progressively and differentiatedly (by equipment type and region) automatically adjust the parameters in the model library (such as factor weights and thresholds), enabling the model to continuously evolve in practice and better reflect actual risks.

[0077] 4. Interactive Presentation Layer:

[0078] PC-based panoramic dashboard: Provides a comprehensive monitoring view for operation and maintenance management personnel at the provincial, municipal, and county levels. The large screen can display an overview of the entire network's health, a ranking of negative devices, alarms for abnormal fluctuations in scores, and a dashboard for the execution of operation and maintenance strategies. Management personnel can drill down to any device to view its detailed profile.

[0079] Mobile Supply Service App: Provides mobile tools for on-site inspection and emergency repair personnel. It receives and processes various work orders dispatched by the system, allows them to view equipment negative lists and handling instructions within the work orders, fill in the handling results on-site, and send photos back, realizing a mobile closed loop for work orders.

[0080] II. Detailed Explanation of Core Algorithm Flow

[0081] To calculate the negative integral of a distribution transformer at time t. For example, its core algorithm flow is as follows:

[0082] 1. Event Monitoring and Factor Matching: The calculation engine monitors the data stream. Suppose that around time t, the engine captures three related events: Event A (a 30-minute power outage of the transformer), Event B (the transformer's A-phase current exceeds the limit for 2 hours), and Event C (a general complaint work order is generated by its customer). The engine immediately matches these three events to the "Outage Factor," "Electrical Anomaly Factor," and "Customer Impact Factor" defined in the model library for this type of transformer, respectively.

[0083] 2. Raw Score Calculation: Based on predefined scoring mapping rules in the model library, raw events are quantified into scores. For example, the rules stipulate: power outage duration of [15, 60) minutes scores 3 points; current exceeding limits duration of [1, 4) hours scores 2 points; each general complaint scores 1 point. Therefore, .

[0084] 3. Time Decay Correction: Considering that the impact of historical events should decrease over time, the algorithm applies a decay function to the score of each event. Assuming event A occurred 1 day ago and event B occurred 0.5 days ago, an exponential decay function is used. Let the attenuation coefficient k = 0.1. Then the score after attenuation is: , Event C is a newly occurring event with a decay coefficient of 1.

[0085] 4. Synergy Effect Judgment and Correction: The algorithm checks whether these events fall within the "synergistic scenario" defined by the model. The model defines a rule: "Power outage event" and "customer complaints occurring in a short period of time" have a synergistic effect. Since events A (power outage) and C (complaint) occur successively within a short period of time, the algorithm determines that synergy is triggered. The synergy correction coefficient is set to 1.3. Therefore, the scores related to events A and C need to be multiplied by this coefficient.

[0086] 5. Weighted Integral Aggregation: Reads the weights of each factor for this device from the model library. Assume the initial weight configuration is as follows: The incremental contribution of this event cluster to the point in time is:

[0087]

[0088]

[0089] 6. Total Score Update and Health Assessment: The increment is added to the device's historical score base value to obtain the latest total score. Subsequently, The current health level is immediately determined by comparing it with the preset health threshold range in the model library (for example, 0-10 is healthy, 10-25 is sub-healthy, 25-40 is of concern, and above 40 is serious).

[0090] 7. Triggering Downstream Applications: Points updates and health assessment results are published in real time. The device health profile service updates the profile of the transformer accordingly. Simultaneously, the intelligent operation and maintenance strategy engine checks whether the new health level meets the triggering conditions of any preset strategies. For example, if this points update causes its health level to move from "sub-healthy" to "attention," a rule is triggered, automatically generating a "diagnostic inspection" work order, which is then pushed to the work order management system and mobile app via a message queue.

[0091] III. Complete Business Closed Loop

[0092] This system realizes a complete business loop from "data perception" to "decision execution" and then to "effect feedback":

[0093] 1. Sensing: Real-time sensing of power grid equipment status and customer feedback through multi-source data access.

[0094] 2. Assessment: Using a negative integral model, the perceived information is quantified, integrated, and assessed to output the device health status.

[0095] 3. Decision-making: Based on the health assessment results, intelligently match operation and maintenance strategies and automatically generate executable tasks.

[0096] 4. Execution: Deliver tasks to on-site personnel via mobile tools and guide them in completing them.

[0097] 5. Feedback: On-site handling results are transmitted back via mobile device. The system deducts equipment points accordingly and feeds the data from the entire chain of "assessment-decision-execution" into the model self-optimization service to optimize future assessments and decisions.

[0098] Through this process, this embodiment transforms scattered data into a unified understanding of equipment health, and automatically translates this understanding into precise operation and maintenance actions, forming an intelligent operation and maintenance system that continuously learns and optimizes, effectively improving the predictability, accuracy, and efficiency of distribution network operation and maintenance.

[0099] Example 2

[0100] Reference Appendix Figure 3 Taking a 10kV Dongcheng Line 5 distribution line as an example, this paper illustrates the entire process of negative point accumulation and operation and maintenance triggering. At 9:00 AM on July 15, the system received three data points within one minute: the distribution automation system reported a ground fault in phase C of the line, which was isolated and restored after 25 minutes; the electricity information collection system showed that the voltage at the monitoring points of the three distribution transformers at the end of the line dropped to 0.8kV (nominal 10kV) during the fault; and the marketing system pushed 95598 repair work orders from eight households in the power supply area of ​​the line.

[0101] The negative integral calculation engine immediately initiates comprehensive calculations. First, it calculates the raw scores for each factor: a 25-minute power outage, based on the mapping rules. Points; if the voltage drop exceeds 15%, according to the rules... 8 related repair requests, according to the rules... Points. Since the three factors trigger simultaneously within a very short time window (within a 1-minute time window), activating the collaborative correction term, assuming... , The collaborative correction coefficient is Assuming all events are newly occurring and the time decay coefficient is 1, the integral increment for this event is: The original score for this route was 15.2 points (sub-healthy state), and after the addition, it reached 19.61 points, still in the "sub-healthy" range but close to the "attention" threshold.

[0102] The equipment health profile service was updated immediately. On the power grid map, the color of the "10kV Dongcheng Line 5" changed from yellow to dark yellow with a slight flicker. Clicking on the line brought up a profile details page showing: a real-time score of 19.61 points, indicating a "sub-healthy" health status; a pie chart showing the score composition indicated that this event contributed 22.5% to the score increase; and the event list allowed drill-down access to three detailed records and their respective contributions. Simultaneously, the intelligent operation and maintenance strategy service, based on the rule of a "sub-healthy" status approaching the "attention" threshold, generated a work order for a "special inspection within 7 days," which was pushed to the line's responsible person, Mr. Zhang, via the power supply service APP. Mr. Zhang conducted an on-site inspection on July 17th, discovered and removed a bird's nest hazard, uploaded photos of the removal in the APP, and closed the work order. Based on the "general hazard elimination" scoring rule, the system deducted 2 points from the line, reducing its score to 17.61 points.

[0103] Example 3

[0104] Reference Appendix Figure 4 Taking the "Huafu Community No. 1 Public Transformer" in operation as an example, this illustrates how the model identifies hidden defects through long-term trends. During the 90-day monitoring period, no power outages occurred at this transformer, but the system recorded the following persistent anomalies: Enterprise-level measurement center data showed that its A-phase current consistently reached 115%-120% of the rated value during the daily evening peak hours (18:00-21:00), exceeding the limit for approximately 3 hours daily; simultaneously, its three-phase current imbalance remained consistently at 25%-30% (national standard requires no more than 15%). Marketing system data showed that there were an average of 3-5 customer inquiries per month regarding "flickering lights" within its power supply area.

[0105] The negative integral calculation engine performs integral calculations daily. (Regarding the current limit exceedance factor...) According to the rules, exceeding the daily limit by more than 2 hours earns 1 point. Because this occurs continuously every day and decays slowly over time, this factor consistently contributed approximately 56 points over 90 days. Regarding the three-phase imbalance factor... Another sub-item, consistently exceeding the limit, is awarded 0.5 points per day, contributing approximately 45 points. (Customer Inquiry Work Order Factor) One point is awarded per month, for a total of 3 points. Although the daily point increase is small, there are no point-deducting events, and the total points show a slow, linear growth. By the 85th day, the total points reach 31.5 points, breaking through the 30-point threshold, and the health status automatically changes from "sub-healthy" to "attention".

[0106] The system then generates a "diagnostic test" recommendation work order and sends it to the relevant work team. The team uses a live-line tester to conduct on-site testing and discovers a 90°C overheating point (normally below 75°C) at the A-phase winding joint of the transformer, constituting a major safety hazard. This successful "early warning-confirmation" case is recorded. After the joint is repaired and replaced, the system, according to the "major safety hazard elimination" rule, deducts 15 points and marks the event pattern for future model optimization.

[0107] Example 4

[0108] Reference Appendix Figure 5 Taking the analysis and optimization of a quarter's operational data by the model self-optimization module as an example, we can illustrate its self-evolution capability. The module collected 12,500 equipment maintenance work order records that occurred and were closed-loop throughout the province this quarter, as well as more than 500,000 related score change events.

[0109] First, the module performs a multi-feature contribution analysis. The analysis reveals that for equipment groups like "heavy-load industrial parks," the combined effect coefficient of the factors "current exceeding limits" and "load fluctuation rate" on the final integral contribution reaches an actual value of 0.35, far exceeding the initial set value of 0.2. However, this combined effect is not significant for the "light-load residential area" group. The analysis also found that for older cables that have been in operation for more than 15 years, the correlation coefficient between the "customer complaints" factor and subsequent actual failures is as high as 0.7, while for newly commissioned cables, this coefficient is only 0.2.

[0110] Next, the module initiates a gradual weight adjustment. Based on the above analysis, it adjusts the model non-parametrically: for the equipment group in the "heavy-load industrial park," the weight of the "current limit violation" factor is adjusted. The coefficient was increased from 0.25 to 0.30, and the synergy coefficient for this group was also adjusted. The weighting for "customer complaints" will be increased from 0.2 to 0.3. For the group of cables with over 15 years of service, the weighting for "customer complaints" will be adjusted. The value was increased from 0.25 to 0.33. Meanwhile, a temporary integral factor for "top oil temperature" was added to all oil-immersed distribution transformers for the high-temperature period in summer, with a weight of 0.1.

[0111] After optimization, the system entered the next observation cycle. Statistical data shows that under the new model parameters, the number of "high-risk" early warning work orders generated by the system decreased by 18% compared to the previous period, but the proportion of work orders confirmed as genuine hidden dangers on-site (i.e., accuracy rate) increased from 68% to 85%. Ineffective inspection workload was significantly reduced. The model self-optimization module recorded the complete chain of this parameter adjustment and effect evaluation in the knowledge base as the basis for the next round of optimization.

[0112] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for health assessment and intelligent operation and maintenance of distribution network equipment based on a negative integral model, characterized in that, include: S1. Model Construction: For lines and distribution transformers in the distribution network, a negative integral calculation model is constructed and maintained respectively. The model defines multiple integral factors, the weight of each factor, the rules for integral accumulation and operation and maintenance score elimination, and the integral threshold corresponding to different health levels. S2. Data-driven point calculation: Real-time access to multi-source data from electricity consumption information collection, power distribution automation, dispatch automation and marketing service systems, and dynamic calculation and updating of negative point values ​​for each device based on the negative point calculation model; S3. Health Assessment and Visualization: Based on the real-time negative score value of the equipment, its health level is determined, and a health profile of the equipment that integrates score details, historical events and ledger information is generated and visualized and alerted on a power grid map. S4. Intelligent Operation and Maintenance Strategy Generation: Based on the health level and profile information of the equipment, automatically trigger or assist in generating differentiated operation and maintenance work orders, maintenance plan suggestions or resource allocation schemes.

2. The method for health assessment and intelligent operation and maintenance of distribution network equipment based on a negative integral model according to claim 1, characterized in that, The integral factors in the negative integral calculation model include at least: equipment downtime frequency and duration factor, equipment electrical measurement anomaly factor, and related customer complaint work order factor; wherein, the weight of the customer complaint work order factor is dynamically adjusted according to the authority of the complaint source channel and the severity of the complaint issue.

3. The method for health assessment and intelligent operation and maintenance of distribution network equipment based on a negative integral model according to claim 2, characterized in that, The method also includes an adaptive optimization step for the integral factor: based on the correlation analysis between the trend of equipment integral changes in the historical period and subsequent failure events, the weight configuration of the integral factor is dynamically optimized using machine learning algorithms.

4. The method for health assessment and intelligent operation and maintenance of distribution network equipment based on a negative integral model according to claim 1, characterized in that, In step S3, generating a device health profile specifically includes: establishing a point contribution tracing mechanism to thoroughly examine the details of the main events that lead to the current negative points; at the same time, integrating the results of device topology consistency verification into the profile to directly associate data anomalies with device health assessment.

5. The method for health assessment and intelligent operation and maintenance of distribution network equipment based on a negative integral model according to claim 1, characterized in that, In step S4, the generation of differentiated operation and maintenance strategies specifically includes: For equipment that reaches the warning threshold, a list of negative issues for that equipment will be automatically added to the planned inspection work order to guide on-site personnel to focus on inspections. In proactive repair or fault reporting scenarios, if the work order involves equipment with high negative scores, the historical anomaly records and handling suggestions for that equipment will be pushed to the repair personnel on the mobile application. The negative points of equipment will be used as an important basis for decision-making in the establishment and reserve of distribution network projects, and priority will be given to the transformation of high-frequency negative equipment.

6. The method for health assessment and intelligent operation and maintenance of distribution network equipment based on a negative integral model according to claim 1, characterized in that, The method also includes a closed-loop feedback mechanism: tracking and recording the processing results of maintenance work orders triggered by negative points, and deducting equipment points according to the point reduction rules if the problem is confirmed to be resolved; at the same time, the processing results are fed back as sample data to the point calculation model to optimize the accuracy of points and point reduction rules.

7. A power distribution network equipment health assessment and intelligent operation and maintenance system for implementing the method of any one of claims 1-6, characterized in that, include: Model Management Center: Provides a graphical interface for configuring and maintaining negative score calculation models for different device types, including defining factors, weights, rules, and thresholds; Real-time integration calculation engine: integrated with the power grid resource business platform, measurement center and marketing system, it acquires data in real time and drives the integration model to perform calculations, outputting dynamic integration of equipment; Equipment Health Profile Service: Receives points data, generates and manages panoramic health profiles of equipment, and provides multi-dimensional query, traceability analysis and visualization services; Intelligent operation and maintenance strategy service: Based on the health information output by the profile service and combined with the pre-set strategy library, it automatically generates or recommends specific operation and maintenance action plans and pushes them to the power supply service command system work order process or mobile application.

8. The system according to claim 7, characterized in that, The system also includes a model self-optimization module, which automatically adjusts the parameters of the negative integral calculation model periodically or triggeredly based on the data sequence of "integral event handling results" accumulated by the operation and maintenance closed-loop feedback mechanism.

9. The system according to claim 8, characterized in that, The model self-optimization module performs the following steps to optimize the negative integral calculation model: a) Multi-feature contribution analysis: Extract and combine features of various events that lead to an increase in equipment score in historical data, and quantify the contribution fluctuation of each score factor to the final score result under different operating scenarios. b) Gradual weight adjustment: Based on the contribution fluctuation obtained in step a), and combined with the effective handling rate of recent maintenance work orders, the weights of each integral factor in the model are adjusted non-linearly and gradually to make the weight allocation more in line with the actual risk level. c) Timing pattern recognition and rule fine-tuning: Identify the periodic abnormal change patterns of device integrals under specific seasons, weather or load modes, and make adaptive fine-tuning of the time decay coefficient or event aggregation conditions in the integral accumulation rules accordingly. d) Grouping and Differentiation Feedback: The equipment is grouped according to type, region, and years of operation. The differences in the correlation between integral changes and failure occurrence among different groups are analyzed, and the differences are fed back to the weight correction process in step b) to achieve group-differentiated adaptation of model parameters. e) Iterative evaluation of strategy value: Establish an evaluation mechanism to retrospectively evaluate the actual effect of operation and maintenance strategies triggered by the optimized model, and use the evaluation results as the target guide for the next round of model parameter optimization.

10. The system according to claim 9, characterized in that, The multi-feature contribution analysis in step a) employs a dynamic feature cross-analysis method. It not only analyzes the independent contribution of a single integral factor, but also focuses on analyzing the synergistic amplification or suppression effect on the integral when two or more factors appear simultaneously under specific spatiotemporal conditions. This effect is quantified into a dynamic synergistic correction coefficient for more refined integral calculation.

Images