Data-driven rural distribution network transformer area reconstruction priority evaluation method and system

Abstract

The application is a rural distribution network transformer area reconstruction priority evaluation method and system based on data driving, which relates to the technical field of distribution network operation and reconstruction, comprising: receiving the output standardized evaluation data set, respectively calculating the distribution transformer load rate, the bare wire proportion, the voltage quality, the power supply reliability and the equipment operation life, five exclusive evaluation indexes adapting to the characteristics of rural distribution network, comparing with the preset threshold value to determine item by item, the indexes not reaching the threshold value are directly assigned as 0, and the indexes reaching the threshold value retain the original quantitative value and output to the downstream. In the application, multi-source heterogeneous distribution network data is fused, the data barriers of different business systems are broken through, the multi-dimensional information of spatial topology, operation state and equipment account is integrated, and comprehensive data support is provided for the evaluation work. The customized evaluation indexes are suitable for the typical operation pain points of rural transformer areas, the threshold value determination mechanism is combined to distinguish the problem states of different transformer areas, and the integrity and pertinence of the evaluation dimensions are ensured.

Description

Technical Field

[0001] This invention relates to the field of power distribution network operation and maintenance technology, and in particular to a data-driven method and system for prioritizing the transformation of rural power distribution transformer areas. Background Technology

[0002] With the continuous advancement of the rural revitalization strategy, the demand for electricity for rural production and daily life is steadily increasing. As the terminal unit of rural power supply, the 0.4kV low-voltage distribution transformer area directly affects the reliability of power supply and the user's electricity experience. Orderly carrying out transformer area upgrades and renovations, and rationally arranging the implementation sequence of upgrades, are routine tasks in distribution network operation and maintenance management, and also important measures to ensure the healthy operation of the rural power grid and improve the level of power supply services.

[0003] Currently, the priority determination for rural power distribution network (PDN) upgrades is largely based on a combination of on-site operation and maintenance experience, operational data statistics, and existing evaluation models. Among these methods, manual on-site inspections are widely used in grassroots PDN management because they can identify problems based on actual on-site conditions. The screening and ranking method based on a single operational indicator is simple to operate and can quickly identify prominent issues. Some regions also adopt mature PDN evaluation systems, conducting comprehensive assessments through weighted scoring. All these methods have played their respective roles in different management scenarios, supporting the orderly implementation of rural PDN upgrades.

[0004] In practical applications, there is still room for improvement in existing evaluation methods; On the one hand, data such as GIS topology, electricity consumption data collection, and equipment ledgers belong to different business systems, resulting in insufficient depth of data integration and application, and the data support dimensions for the assessment process need to be further enriched. On the other hand, rural transformer substations are characterized by a high proportion of bare conductors, large seasonal fluctuations in load, and long equipment service cycles, and there is still room for improvement in the adaptability of indicators and the pertinence of quantitative methods in some general assessment systems. In addition, most assessments are based on static, single-time judgments, and there is still room for improvement in dynamically adapting to the evolution of equipment status and supporting closed-loop management throughout the entire transformation cycle. Summary of the Invention

[0005] The purpose of this invention is to propose a data-driven method and system for prioritizing the renovation of rural power distribution network areas in order to solve the above-mentioned problems.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: A data-driven priority assessment system for rural power distribution network area renovation includes: The multi-source data access and preprocessing module connects to three heterogeneous data sources: GIS topology, electricity consumption data collection, and PMS equipment ledger. It automatically completes data deduplication, outlier removal, missing value completion, and format standardization. It aggregates data according to the unique identifier of each transformer area to form a standardized evaluation dataset of "one file per transformer area". It also adopts a two-layer architecture of persistent storage and memory caching to complete data management. The core indicator calculation and threshold determination module receives the output standardized evaluation dataset and calculates five exclusive evaluation indicators adapted to the characteristics of rural power distribution networks: distribution transformer load rate, bare conductor ratio, voltage quality, power supply reliability, and equipment operating years. It then judges each indicator against the preset threshold. Indicators that do not reach the threshold are directly assigned a value of 0, while indicators that reach the threshold retain their original quantitative values ​​and are output downstream. The core indicator normalization processing module receives the values ​​of each indicator after threshold determination, linearly maps non-zero indicators to standard scores of 0-100 points according to the relative ranking of the transformer area, and directly retains 0 points for zero-value indicators, and outputs the normalized scores of each individual indicator of all transformer areas. The weighting configuration and comprehensive score calculation module receives the normalized scores of each individual indicator, performs weighted summation according to the differentiated weighting system prioritizing safety and people's livelihood, and obtains the comprehensive evaluation score of the transformer area. The priority sorting output and dynamic management module receives the comprehensive evaluation score, generates a priority list of transformer area renovations in descending order of score, supports visual display of results and standardized report export, and also has the ability to dynamically re-evaluate and trace historical evaluation files.

[0007] Preferably, the multi-source data access and preprocessing module has a built-in multi-level data cleaning and governance unit, which sequentially performs deduplication, outlier removal, missing value completion, and cross-source logical consistency verification; when connecting to GIS topology data, it synchronously verifies the spatial topology attribution relationship of transformer area-line-user; when connecting to electricity consumption collection data, it synchronizes the operation data of the past 12 months on a daily basis and verifies the data integrity; when connecting to PMS ledger data, it updates static equipment information on a quarterly basis and verifies the mapping relationship of transformer area-distribution transformer; and adopts a two-layer storage architecture of persistent storage of full data in relational database and caching current evaluation data in memory database.

[0008] Preferably, in the core indicator calculation and threshold determination module, the distribution transformer load rate is taken as the maximum daily load rate of the past 12 months, and the threshold is set to 85%, with a value of 0 if it is lower than the threshold; the bare conductor ratio is the proportion of the length of bare conductors in the transformer area to the total length of low-voltage lines, with a value of 0 if the ratio is 0; the voltage quality is determined by the user-side voltage being lower than 198V as the low voltage judgment standard, and the low voltage occurrence rate of the past 12 months is statistically analyzed, with a value of 0 if the occurrence rate is 0; the power supply reliability is statistically analyzed by the total number of unplanned power outages in the past 12 months, with the original values ​​retained for all transformer areas; the equipment operating life is calculated based on the years of operation of the distribution transformer, with a threshold of 20 years, and a value of 0 if it is lower than the threshold.

[0009] Preferably, the indicator normalization processing module adopts a zero-value isolation relative ranking normalization method: for each indicator, non-zero value transformer areas are first selected to form a ranking pool for evaluation, and zero-value transformer areas are directly marked with a score of 0 for that indicator; the transformer areas in the ranking pool are arranged in ascending order of indicator value, and the ranking system is used to process tied values. The score of a single indicator = the current ranking position of the transformer area / the total number of participating transformer areas × 100, and the score range is 0~100 points. The more serious the problem, the higher the score.

[0010] Preferably, the differentiated weight system of the weight configuration and comprehensive score calculation module is as follows: bare conductor ratio 25%, voltage quality 25%, power supply reliability 20%, distribution transformer load rate 15%, and equipment service life 15%; the comprehensive evaluation score is the sum of the products of the normalized scores of the five indicators and their corresponding weights, and the higher the score, the higher the priority of the transformation; the system supports the storage of multiple weight schemes and one-click switching to adapt to different special transformation evaluation scenarios.

[0011] Preferably, the priority sorting output and dynamic control module adopts a multi-level sorting rule: the core sorting basis is the comprehensive score in descending order. When the comprehensive scores are the same, the secondary sorting is carried out in turn according to the bare conductor ratio score, voltage quality score, and power supply reliability score. It supports multi-dimensional filtering and sorting by power supply station, administrative region, and problem type, and can generate a special transformation priority list for the corresponding dimension.

[0012] Preferably, the priority ranking output and dynamic management module has a dual dynamic update mechanism of periodic iteration and event-triggered iteration: it automatically pulls the latest 12 months of operating data at a fixed annual cycle to complete the annual re-evaluation of all transformer areas and update the annual renovation priority list; when a transformer area completes renovation acceptance, experiences a major fault, or the ledger information is updated, it supports triggering batch real-time re-evaluation of a single transformer area or a designated area and adjusting the priority ranking in real time.

[0013] Preferably, the priority sorting output and dynamic management module supports dual-format display of list details and GIS spatial heat map: the list displays basic information of the transformer area, original indicators, individual scores, comprehensive scores and problem tags; the GIS map marks the priority level of the transformer area with different colors, intuitively presenting the spatial distribution of high-risk transformer areas; it supports exporting standardized Excel reports containing complete assessment fields, directly supporting the renovation project initiation and fund allocation work.

[0014] Preferably, the system adopts a modular loosely coupled architecture, reserving interfaces for index expansion, external system integration, and algorithm upgrades; it supports the addition of assessment indicators such as three-phase imbalance and low-voltage line loss rate, and can be integrated with project management systems and operation and maintenance control systems to achieve assessment-project initiation-implementation-post-assessment management, while being compatible with future business expansion needs for rural power distribution network management.

[0015] A data-driven method for prioritizing the upgrading of rural power distribution transformer areas includes: Step 1 involves multi-source data access and preprocessing, automatically connecting to three heterogeneous data sources: GIS topology, electricity consumption data collection, and PMS equipment ledger. After multi-level cleaning and treatment, the data is deduplicated, outliers are removed, missing values ​​are filled, and cross-source logic is verified. The data is aggregated according to the unique identifier of the transformer area to form a standardized evaluation dataset. A dual-layer storage architecture is adopted to balance full data traceability and batch calculation efficiency. Step two involves calculating core indicators and determining thresholds. Five rural power distribution network-specific evaluation indicators are calculated in batches based on a standardized dataset. Each indicator is then judged against the preset thresholds. Indicators without corresponding hidden dangers are directly assigned a value of 0, while indicators with problems retain their original quantitative values ​​and proceed to the next step. Step 3 is the indicator normalization process, which involves isolating individual indicators at zero values ​​and arranging non-zero indicators in ascending order of value. The relative ranking is linearly mapped to a standard score of 0 to 100, eliminating the difference in units and achieving fair comparison across different regions. Step four involves weight configuration and comprehensive score calculation. Differentiated weights are configured in accordance with the principle of prioritizing safety and people's livelihood. The scores of the five indicators are normalized and weighted to obtain the comprehensive evaluation score. Multiple weight schemes can be switched to adapt to different scenarios. Step five involves priority sorting and dynamic management. A priority list is generated in descending order of comprehensive score, supporting multiple display formats and report export. It also has dynamic iteration capabilities for both periodic and real-time re-evaluation and retains historical archives to support effect traceability.

[0016] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are: 1. This invention integrates multi-source heterogeneous distribution network data, breaking down data barriers between different business systems and consolidating multi-dimensional information such as spatial topology, operational status, and equipment ledgers to provide comprehensive data support for assessment work. Customized assessment indicators are tailored to the typical operational pain points of rural distribution areas, and a threshold judgment mechanism distinguishes the problem status of different distribution areas, ensuring the completeness and relevance of the assessment dimensions.

[0017] 2. This invention replaces the traditional manual statistics, calculation and tabulation work mode with a fully automated evaluation process, which can support batch evaluation of large-scale transformer areas, effectively shorten the evaluation cycle and reduce the transactional workload of grassroots operation and maintenance personnel; the modular loosely coupled architecture has flexible parameter configuration capabilities, and can adjust the evaluation rules according to the operation characteristics and management needs of different regions, adapting to diverse special transformation evaluation scenarios. Attached Figure Description

[0018] Further details, features, and advantages of this application are disclosed in the following description of exemplary embodiments in conjunction with the accompanying drawings, in which: Figure 1This is a flowchart of the method of the present invention. Detailed Implementation

[0019] Several embodiments of this application will now be described in more detail with reference to the accompanying drawings to enable those skilled in the art to implement this application. This application may be embodied in many different forms and for various purposes and should not be limited to the embodiments set forth herein. These embodiments are provided to make this application thorough and complete, and to fully convey the scope of this application to those skilled in the art. The embodiments described do not limit this application.

[0020] Unless otherwise defined, all terms used herein (including technical and scientific terms) shall have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. It will be further understood that terms such as those defined in commonly used dictionaries shall be interpreted as having a meaning consistent with their meaning in the relevant field and / or the context of this specification, and shall not be interpreted in an idealized or overly formal sense unless expressly defined herein.

[0021] Example 1 Its specific implementation method is combined with the appendix Figure 1 Please provide a detailed explanation.

[0022] In this embodiment, it includes: Module 1: Multi-source data access and preprocessing module This module consists of five units: GIS topology data docking unit, electricity consumption acquisition system data docking unit, PMS equipment ledger docking unit, multi-level data cleaning and verification unit, and dual-layer data storage and caching unit. The data cleaning and verification unit has four sub-steps: deduplication, outlier removal, missing value completion, and logical consistency verification, which realizes full-link data quality control.

[0023] As the data entry layer of the entire assessment system, this module breaks down data barriers between three heterogeneous production systems—GIS, electricity consumption data collection, and PMS—through standardized business interfaces. After retrieving raw data by unique identifier of each transformer substation, it first unifies the format and standardizes the fields. Then, it undergoes multi-level quality verification and cleaning to remove invalid, abnormal, and erroneous data. Finally, it aggregates data by transformer substation dimension to form a standardized assessment dataset of "one file per transformer substation." It adopts a two-layer architecture of "persistent storage + memory caching." Full data is stored locally to ensure traceability, while current assessment data is cached in memory to improve computational efficiency, providing high-quality and highly available basic data support for subsequent full-process indicator calculations.

[0024] For multi-source data precise integration with GIS topology data, a spatial data service interface is used. The extracted core data includes spatial polygons of the distribution area's power supply range, low-voltage line vector data, conductor type / diameter / length attributes, tower coordinates and parameters, spatial markers for tree-blocked sections, markers for river and road crossing sections, and spatial locations of user meters. After integration, a topology consistency check is automatically performed to ensure that each low-voltage line belongs to a unique distribution area and each user location falls within the corresponding distribution area's power supply range. Data with abnormal topology attribution is automatically marked and pushed for manual review, preventing spatial data misalignment from the source. For electricity consumption data integration, a standardized business interface is used to synchronize distribution area-level operational data daily, extracting extreme values ​​of total active power, reactive power, three-phase voltage, fault trip records, and unplanned power outage event records for the past 12 months. Data granularity supports 15-minute sampling and can be aggregated into daily and monthly statistics as needed. Data integrity is checked synchronously during integration; if the monthly data missing rate for a single distribution area exceeds 10%, it is marked as unqualified and temporarily excluded from the current evaluation. The PMS equipment ledger data is integrated using a database view synchronization method, updating static ledger information quarterly. This involves extracting data such as transformer rated capacity, rated power factor, equipment model, commissioning date, previous modification records, pole specifications and quantity, and equipment maintenance and repair files. After integration, the mapping relationship between equipment and distribution areas is verified to ensure that each distribution area corresponds to at least one main transformer. Distribution areas without main equipment are marked as abnormal, and maintenance personnel are notified to complete the mapping.

[0025] The multi-level data cleaning and governance process involves three levels. The first level is deduplication, which uses a unique transformer substation code and a timestamp for deduplication. For duplicate data at the same time point, only the most recently entered valid record is retained to prevent errors in subsequent indicator calculations. The second level is outlier removal, which uses dual thresholds based on both physical reasonable ranges and business logic. For example, the reasonable range for transformer load rate is set to 0-120%, and the reasonable range for single-phase voltage is set to 160V-260V. Values ​​outside these ranges are directly identified as outliers and removed. Substations with a load rate of 0 for several consecutive days are considered out of service or have abnormal data and are excluded from the evaluation scope. The third level is missing value completion. For short-term operational data with no more than 7 consecutive days of missing data, linear interpolation of adjacent dates within the same substation is used for completion. For missing ledger fields, priority is given to completing them through association with other data sources; fields that cannot be completed are marked as missing, and the corresponding indicators are temporarily excluded from calculation. The fourth level is logical consistency verification, which verifies the logical rationality of data across data sources. For example, if the total length of the transformer area lines in the GIS statistics deviates from the line length recorded in the PMS ledger by more than 15%, an alarm will be automatically triggered to push for manual verification, so as to avoid the distortion of the subsequent overall assessment due to errors in the basic data.

[0026] The dual-layer storage architecture uses a relational database to store all historical data, including raw access data, intermediate calculation data for each round of evaluation, and result data. All operations are traceable, ensuring data traceability throughout its entire lifecycle. An in-memory cache database stores the basic dataset and intermediate calculation results for the current evaluation, significantly improving the data retrieval efficiency for batch calculations of indicators. At a scale of thousands of distribution areas, data retrieval time can be reduced by more than 90%.

[0027] Module Two: Core Indicator Calculation and Threshold Determination This module contains six units: transformer load rate calculation unit, bare conductor ratio calculation unit, voltage quality statistics unit, power supply reliability calculation unit, equipment service life calculation unit, and threshold rule determination unit. The first five units correspond to the exclusive calculation logic of a core indicator, while the threshold determination unit is a common execution unit that uniformly completes the qualification determination and assignment of all indicators.

[0028] Using a single transformer substation as the smallest calculation unit, the pre-processed corresponding data source is invoked to quantify five indicators according to the calculation formula specific to rural power distribution networks. The calculation results for each indicator are then input into a threshold determination unit and compared with preset thresholds specific to rural transformer substations. Substations that do not reach the threshold indicate no urgent need for upgrading in that dimension, and the corresponding indicator is directly assigned a value of 0 and not included in subsequent sorting. Substations that reach the threshold retain their original quantified values ​​and proceed to the next stage of normalization processing. The entire calculation process supports batch parallel execution, enabling the calculation of indicators for all transformer substations in the entire region at once.

[0029] The calculation logic for the transformer load rate index is as follows: the maximum daily load rate within the past 12 months is selected as the final evaluation value. The daily load rate is calculated by dividing the maximum daily total active power by the product of the transformer's rated capacity and rated power factor, and then multiplying by 100%. During the calculation process, abnormally high load data during equipment maintenance, testing, and commissioning are automatically removed to avoid interference from non-realistic operating data in the evaluation results. If a distribution area contains multiple transformers, the load rate of each transformer is calculated separately, and the maximum value is taken as the overall load rate of the area to ensure that the most severe overload scenarios are captured. The threshold rule uses 85% as the overload judgment threshold, which is set with reference to the overload standard for rural distribution transformers in the distribution network operation regulations. Distribution areas with a load rate greater than or equal to 85% are judged as overloaded and overloaded, and the original load rate value is retained for subsequent sorting; distribution areas with a load rate below 85% are judged as having normal load, and this index is assigned a value of 0. Using the annual maximum value instead of the monthly average value as the assessment basis is mainly suitable for the characteristics of large seasonal fluctuations in rural load. It can accurately cover extreme scenarios such as concentrated electricity consumption during the Spring Festival return home and sudden load increases in agricultural irrigation, avoiding the problem of missing the assessment of phased heavy loads.

[0030] The calculation logic for the bare conductor percentage index is based on GIS topology data. First, the total length of 0.4kV low-voltage lines within the power grid's ownership area is calculated, including main lines and branch lines, but excluding service connections owned by users. Then, the cumulative length of overhead bare conductors is calculated. The final bare conductor percentage is equal to the cumulative length of bare conductors divided by the total length of low-voltage lines, multiplied by 100%. The calculation automatically distinguishes between insulated conductors, overhead bare conductors, and buried cables, including only overhead bare conductors in the hazard statistics. A threshold of 0 is used as the judgment threshold. If bare conductors are present in a transformer area, there is a safety hazard such as electric shock or tree discharge, and the original percentage value is retained for subsequent ranking. For transformer areas with a bare conductor percentage of 0, this index is assigned a value of 0. This threshold setting is tailored to the actual operation of rural distribution networks. Rural areas have weaker control over human activities and more trees and vegetation below the lines; even a small number of bare conductors pose a safety risk, therefore, they are included in the assessment as long as they exist.

[0031] The voltage quality index calculation logic is based on user-side voltage data from the power consumption acquisition system, statistically analyzing low-voltage occurrences over the past 12 months. Low voltage is determined according to national power quality standards: a single-phase user-end voltage below 90% of the rated voltage (198V) is considered low voltage. To avoid misjudgments caused by instantaneous voltage fluctuations, a rule is adopted where only three consecutive sampling points below the threshold are counted as one valid low-voltage point. The final low-voltage occurrence rate equals the total number of valid low-voltage points divided by the total number of sampling points within the statistical period, multiplied by 100%. The statistics cover all end-user sampling points within the transformer area, focusing on users at the end of the power supply radius to accurately reflect voltage drop issues in long-line power supply. The threshold rule is: transformer areas with a low-voltage occurrence rate greater than 0 are considered to have voltage quality problems, and the original occurrence rate value is retained for subsequent sorting; transformer areas with an occurrence rate of 0 have this index assigned a value of 0.

[0032] The power supply reliability index is calculated by counting the total number of unplanned power outages in the distribution area over the past 12 months. This includes the number of fault trips and temporary outages, excluding planned outages such as scheduled maintenance, business expansion construction, and power supply guarantees for major government events. The statistics are double-checked using switch tripping records and outage event ledgers to avoid double counting or omissions. The final index value is the sum of the two types of unplanned outages; a higher number indicates more frequent faults and poorer power supply reliability in the distribution area. The threshold rule is that no zero threshold is set for this index; all raw counts from all distribution areas are retained for ranking, even those with zero outages, enabling a horizontal comparison of reliability across all distribution areas.

[0033] The calculation logic for the equipment service life index is as follows: the distribution transformer is used as the representative of the core equipment in the distribution area, and the service life is equal to the current assessment year minus the year the distribution transformer was put into operation. If the distribution transformer has undergone a complete overhaul or capacity expansion and there is an official record, the service life can be recalculated based on the year the expansion was completed. If there are multiple distribution transformers in the distribution area, the service life of the transformer with the longest service life is taken as the overall index value for the distribution area to ensure coverage of the oldest equipment potential hazards. The threshold rule uses 20 years as the threshold for judging old equipment, which corresponds to the conventional design service life of distribution transformers. Distribution areas with a service life of 20 years or more are judged as having equipment that has exceeded its service life and retain the original service life value for subsequent ranking; distribution areas with a service life of less than 20 years are judged as having normal equipment condition, and this index is assigned a value of 0.

[0034] The threshold configurable mechanism allows all threshold parameters to be configured visually in the background. Administrators can adjust the threshold level according to the actual operation level of the local rural power distribution network and the scale of renovation funds to adapt to the differentiated needs of different assessment scopes such as counties and cities. After the parameter is adjusted, no code modification is required, and it takes effect immediately.

[0035] Module 3: Indicator Normalization Processing Module This module consists of four parts: the participating area screening unit, the indicator ranking calculation unit, the score normalization conversion unit, and the zero value isolation marking unit. Each unit is executed sequentially according to the process to complete the conversion from the original indicator value to the standard score of 0 to 100. The five indicators execute the normalization process independently without interfering with each other.

[0036] For each individual indicator, the participating transformer area screening unit first separates transformer areas into those with zero and non-zero values. Transformers with zero values ​​are directly marked with 0 points, while transformer areas with non-zero values ​​enter the ranking pool. The ranking calculation unit sorts the indicators in ascending order of their original values, with lighter problems ranking higher and more severe problems ranking lower. The score conversion unit converts the ranking of the transformer areas into standard scores between 0 and 100 using a linear mapping formula. Finally, the standard scores of all five indicators for all transformer areas are output, providing a unified standard data for subsequent weighted calculations.

[0037] The selection of participating transformer substations and the zero-value isolation process are as follows: For each indicator, substations with indicator values ​​greater than 0 are selected from all substations to be evaluated to form a ranking pool for that indicator; substations with indicator values ​​of 0 are not included in the ranking pool and are directly marked with a score of 0. The design logic of using zero-value isolation is to avoid a large number of normal substations lowering the ranking differentiation. If all substations with zero values ​​are included in the ranking, the vast majority of non-problematic substations will be concentrated in the extremely low score range, which will compress the score difference between substations with problems and make it impossible to accurately distinguish the severity. After isolation, the ranking pool consists only of substations with the corresponding problems, and the scores can fully reflect the differences in the severity of the problems, significantly improving the evaluation differentiation.

[0038] The ranking rules for indicators rank the stations entering the ranking pool in ascending order based on their original indicator values. Smaller values ​​indicate milder problems and lower rankings, while larger values ​​indicate more severe problems and higher rankings. For stations with identical indicator values, a ranking system is used: if n stations have the same value and are tied for the kth position, the next station with a different value will be ranked k+n. This ensures that stations with the same score have identical scores, and the total ranking corresponds to the total number of participating stations, with the score conversion logic unaffected by ties.

[0039] The normalized score conversion uses a relative ranking linear conversion formula. The final score for a single indicator equals the current district's ranking divided by the total number of districts in the ranking pool, then multiplied by 100. According to this rule, the district with the worst-ranked problem receives 100 points, while the district with the least problem receives close to 0 points, achieving a positive correlation between problem severity and score. Compared to the linear normalization method based on absolute values, this relative ranking method has significant advantages: absolute normalization is easily affected by extreme outliers. For example, if a district has an abnormally high load rate due to special circumstances, it will cause the scores of all other districts to be compressed into a low score range, resulting in extremely poor differentiation. In contrast, the relative ranking method only focuses on the relative position of a district within the same batch, and is not affected by extreme fluctuations in absolute values. The evaluation results are more stable and fairer, and it is especially suitable for horizontal comparison and ranking of hundreds or thousands of districts.

[0040] The batch parallel computing mechanism ensures that the normalization calculations of the five indicators are independent. The system adopts a multi-threaded parallel computing architecture, which can simultaneously execute the screening, ranking, and conversion processes of the five indicators, significantly improving computing efficiency. For transformer substations with a scale of thousands, the normalization calculation of all indicators can be completed within seconds, supporting the need for high-frequency evaluation iterations.

[0041] Module 4: Weighting and Overall Score Calculation This module consists of three parts: a weight parameter configuration unit, a weighted comprehensive score calculation unit, and a score rationality verification unit. The weight configuration unit is responsible for managing the weight parameters and storing multiple schemes, the calculation unit is responsible for performing weighted summation operations, and the verification unit is responsible for multi-level verification to ensure the accuracy of the calculation results.

[0042] Based on the core principles of "safety first, people's livelihood as the priority, and consideration of operation and equipment status" in rural power distribution network renovation, a differentiated fixed weight system is preset. During calculation, the scores of five indicators for each transformer substation are normalized, multiplied by the corresponding weight coefficients, and then summed to obtain the comprehensive evaluation score of the substation. The system supports flexible adjustment of weight parameters and storage of multiple schemes, which can adapt to different evaluation focuses at different times and for different projects. The higher the comprehensive score, the more urgent the need for substation renovation.

[0043] The differentiated weighting system and its design logic: The total weight of the entire weighting system is 100%, with bare conductor ratio and voltage quality each accounting for 25%, power supply reliability accounting for 20%, and transformer load rate and equipment service life each accounting for 15%. The core logic of the weighting setting is to stratify according to the nature of the hidden danger and the urgency of the renovation: The first layer is safety and livelihood indicators, including bare conductor ratio and voltage quality. Bare conductors are directly related to personal safety and the risk of fire in the power grid, falling within the safety red line and being the primary control target for rural distribution network renovation; low voltage at the end directly affects the daily production and life of farmers, representing the most concentrated user demands and the highest complaint rate among livelihood issues. Both have the highest urgency, therefore they are given the highest weight, accounting for half of the total weight, fully reflecting the principle of tilting renovation resources towards safety hazards and livelihood demands. The second layer is operational reliability indicators, namely power supply reliability. Frequent power outages not only affect the quality of power supply services but also reflect the large number of potential faults in the overall equipment of the transformer area and the heavy maintenance pressure. Its importance is second only to safety and livelihood, therefore it is given a weight of 20%. The third layer consists of equipment status indicators, including transformer load rate and equipment service life. Heavy load and overload problems can be temporarily alleviated through three-phase load adjustment, temporary capacity expansion, load transfer and other operation and maintenance methods. Old equipment can still operate stably as long as no faults occur. Both belong to medium and long-term renovation needs and have relatively low urgency. Therefore, each is assigned a weight of 15%.

[0044] The comprehensive score calculation rules for a transformer substation are as follows: The substation load rate score is multiplied by 15%, the bare conductor percentage score by 25%, the voltage quality score by 25%, the power supply reliability score by 20%, and the equipment service life score by 15%. The comprehensive score ranges from 0 to 100. A higher score indicates a more severe comprehensive problem in the substation, requiring a higher priority for improvement. During the calculation, the system automatically controls numerical precision, retaining two decimal places to avoid ranking errors caused by precision mistakes. If a substation is excluded from the calculation due to missing data for a particular indicator, the system will proportionally adjust the effective weights of the other four indicators to ensure the comprehensive score remains within the reasonable range of 0 to 100. Simultaneously, the system will mark the substation as having incomplete data to alert management personnel.

[0045] The flexible weight configuration and multi-scheme management system includes a visual weight configuration interface, allowing administrators to adjust the weight values ​​of various indicators based on annual renovation priorities. For example, when conducting a special campaign to address safety hazards, the weight of the bare conductor percentage can be increased to 35%, while the weights of other indicators can be correspondingly decreased. When conducting a special campaign to address low voltage, the weight of the voltage quality indicator can be increased. The system supports saving multiple weight schemes, each containing independent weight parameters and corresponding evaluation results. Different schemes can be switched with a single click to recalculate the comprehensive score, eliminating the need to repeatedly execute indicator calculation and normalization processes, thus efficiently adapting to different evaluation scenarios.

[0046] The scoring rationality verification system incorporates multi-layered verification rules to ensure calculation accuracy: First, it verifies whether the individual score is within the range of 0 to 100; if it exceeds this range, it backtracks the normalization calculation process. Second, it verifies whether the overall score is within the range of 0 to 100; if it exceeds this range, it checks the weight configuration and weighted calculation logic. Third, it performs sampling verification on the top 10% and bottom 10% of the stations based on the overall score, automatically comparing the original indicators with the scoring logic to identify data anomalies or calculation errors, ensuring the accuracy and credibility of the final overall score.

[0047] Module 5: Priority Sorting Output and Dynamic Management Module This module comprises five parts: a priority sorting unit, a multi-dimensional result display unit, a standardized report export unit, a dynamic iteration control unit, and a historical archive management unit. It is responsible for both the output and application of the current evaluation results and the long-term dynamic updates and archive management. It serves as the exit link connecting the evaluation model with actual business operations.

[0048] Based on the comprehensive score of each transformer substation, the system prioritizes all substations according to a pre-defined multi-level sorting rule, generating a final priority list. The evaluation results are displayed through various formats, including visual pages, GIS maps, and statistical reports. Standardized report export is supported, directly assisting in project initiation, funding allocation, and planning. A dynamic iteration mechanism is also established to periodically update operational data for reassessment, supporting immediate reassessment triggered by events, and retaining all historical evaluation files to achieve dynamic management and full-cycle traceability of renovation priorities.

[0049] The core sorting rule of the multi-level priority ranking system is to arrange the comprehensive scores in descending order, with higher scores ranking higher and having higher priority for renovation. For transformer substations with identical comprehensive scores, a three-level tie-breaking rule is set up, in order of priority: the first level compares the score of the bare conductor ratio, with higher scores ranking higher and prioritizing substations with prominent safety hazards; the second level compares the score of the voltage quality, with higher scores ranking higher and prioritizing substations with strong public demand; the third level compares the score of the power supply reliability, with higher scores ranking higher and prioritizing substations with frequent faults. Through the multi-level ranking rule, the randomness of the ranking of substations with the same score is completely avoided, always adhering to the principle of prioritizing safety and public welfare in renovations. At the same time, the system supports multi-dimensional filtering and sorting, and can filter substations by power supply station, administrative village / town, problem type, etc., generating separate priority lists for special projects such as bare conductor projects, low voltage projects, and old equipment projects, adapting to different types of special renovation project requirements.

[0050] The multi-dimensional results display system offers two display formats: One is a list-style detailed display, which shows basic information about the transformer substations in order of priority, including the power supply station, substation number, substation name, transformer capacity, and the original values, individual scores, overall scores, and evaluation date of five indicators. It also automatically generates problem tags for each substation, enabling rapid identification of core issues. Clicking on a single substation opens a details page, visually presenting the score composition of the five indicators through a radar chart, and automatically generating problem diagnosis and improvement direction suggestions to assist management decision-making. The other format is a GIS spatial visualization display, which uses different colors to mark the priority levels of substations on an electronic map: high-scoring, high-priority substations are marked in red, medium-priority in yellow, and low-priority in green, forming a heat map of improvement priority. This visually presents the spatial clustering characteristics of high-risk substations, facilitating the overall planning of regional improvement projects, achieving contiguous improvement, and improving the efficiency of fund utilization.

[0051] The standardized report export function supports exporting reports in multiple formats and dimensions. The core is a standardized evaluation report in Excel format, which by default includes complete fields such as ranking, affiliated power supply station, transformer area number, transformer area name, transformer capacity, original values ​​of five indicators, normalized scores of the five indicators, comprehensive score, problem type label, and evaluation date. After export, it can be directly used for renovation project applications and funding allocation plan preparation. Custom export is also supported: users can choose to export evaluation lists by administrative region, export special renovation lists by problem type, or extract a project suggestion list from the top N transformer areas based on the annual renovation funding scale. Custom selection of export fields is also supported to adapt to the report format requirements of different management departments, reducing manual processing workload.

[0052] The dynamic iterative management mechanism establishes a dynamic update mechanism that combines regular iteration with event-triggered iteration to ensure the timeliness and accuracy of the priority list. Regular iteration is an annual automatic reassessment; at the end of December each year, the system automatically retrieves the latest operational and ledger data from the past 12 months, automatically completes the entire process of calculation and sorting, and generates the next year's upgrade priority list, adapting to the development trend of increasing rural load and aging equipment year by year. Event-triggered iteration is a manual, real-time reassessment; when a transformer substation completes an upgrade project and passes acceptance, a major power outage causes a change in equipment status, or significant updates occur to equipment ledger information, managers can trigger a batch reassessment of a single substation or a designated area, updating the substation priority ranking in real time. This prevents completed substations from occupying high-priority slots for extended periods, ensuring that upgrade resources are accurately allocated to substations that truly need them.

[0053] The historical archive management and effect tracking system retains complete files for each assessment, including original basic data, threshold and weight parameters, individual indicator scores, comprehensive scores, and priority lists. All files are archived and stored according to assessment batch and assessment time, supporting the ability to review and query assessment results for any historical batch at any time. It also provides a comparative analysis function for the transformation effect, allowing comparison of score changes and ranking fluctuations of the same distribution area in two assessments, automatically analyzing the implementation effect of transformation measures; it can also compare the overall score trend of the region to assess the improvement in the health level of the regional distribution network, providing data support for subsequent optimization of transformation strategies and adjustment of funding directions.

[0054] The scalable architecture design system adopts a modular, loosely coupled architecture, reserving multiple types of extension interfaces: First, an indicator extension interface, which allows for the addition of assessment indicators such as three-phase imbalance, residual current device commissioning rate, and low-voltage line loss rate according to business needs, without requiring a system architecture refactoring; second, a system interface, which can connect to project management systems and operation and maintenance control systems, enabling the assessment results to be directly pushed to the project initiation and implementation processes, forming a closed-loop management of assessment-project initiation-implementation-post-assessment; and third, an algorithm extension interface, which allows for the introduction of more complex intelligent assessment algorithms, upgrading the assessment model without altering the existing processes, and fully adapting to the future business development needs of rural power distribution network management.

[0055] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

[0056] It should be noted that, in this document, the use of relational terms such as "first" and "second" is merely for distinguishing one entity or operation from another, and does not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0057] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0058] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0059] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A data-driven priority evaluation system for rural power distribution network area renovation, characterized in that, include: The multi-source data access and preprocessing module connects to three heterogeneous data sources: GIS topology, electricity consumption data collection, and PMS equipment ledger. It automatically completes data deduplication, outlier removal, missing value completion, and format standardization. It aggregates data according to the unique identifier of each transformer area to form a standardized evaluation dataset of "one file per transformer area". It also adopts a two-layer architecture of persistent storage and memory caching to complete data management. The core indicator calculation and threshold determination module receives the output standardized evaluation dataset and calculates five exclusive evaluation indicators adapted to the characteristics of rural power distribution networks: distribution transformer load rate, bare conductor ratio, voltage quality, power supply reliability, and equipment operating years. It then judges each indicator against the preset threshold. Indicators that do not reach the threshold are directly assigned a value of 0, while indicators that reach the threshold retain their original quantitative values ​​and are output downstream. The core indicator normalization processing module receives the values ​​of each indicator after threshold determination, linearly maps non-zero indicators to standard scores of 0-100 points according to the relative ranking of the transformer area, and directly retains 0 points for zero-value indicators, and outputs the normalized scores of each individual indicator of all transformer areas. The weighting configuration and comprehensive score calculation module receives the normalized scores of each individual indicator, performs weighted summation according to the differentiated weighting system prioritizing safety and people's livelihood, and obtains the comprehensive evaluation score of the transformer area. The priority sorting output and dynamic management module receives the comprehensive evaluation score, generates a priority list of transformer area renovations in descending order of score, supports visual display of results and standardized report export, and also has the ability to dynamically re-evaluate and trace historical evaluation files.

2. The data-driven rural power distribution network substation renovation priority evaluation system according to claim 1, characterized in that, The multi-source data access and preprocessing module incorporates multi-level data cleaning and governance units, sequentially performing deduplication, outlier removal, missing value completion, and cross-source logical consistency verification. When connecting to GIS topology data, it synchronously verifies the spatial topology attribution relationship between distribution areas, lines, and users. When connecting to electricity consumption collection data, it synchronizes the operational data of the past 12 months daily and verifies data integrity. When connecting to PMS ledger data, it updates static equipment information quarterly and verifies the mapping relationship between distribution areas and transformers. It adopts a two-layer storage architecture that persistently stores all data in a relational database and caches current evaluation data in an in-memory database.

3. The data-driven rural power distribution network area renovation priority evaluation system according to claim 1, characterized in that, In the core indicator calculation and threshold determination module, the distribution transformer load rate is taken as the maximum daily load rate of the past 12 months, and the threshold is set to 85%. If it is lower than the threshold, it is assigned a value of 0. The bare conductor ratio is the proportion of the length of bare conductors in the transformer area to the total length of low-voltage lines. If the ratio is 0, it is assigned a value of 0. The voltage quality is judged as the low voltage standard when the voltage on the user side is lower than 198V. The low voltage occurrence rate of the past 12 months is calculated. If the occurrence rate is 0, it is assigned a value of 0. The power supply reliability is calculated as the total number of unplanned power outages in the past 12 months. The original values ​​are retained for all transformer areas. The equipment service life is calculated based on the service life of the distribution transformer. The threshold is set to 20 years. If it is lower than the threshold, it is assigned a value of 0.

4. The data-driven rural power distribution network substation renovation priority evaluation system according to claim 1, characterized in that, The indicator normalization module adopts a relative ranking normalization method with zero-value isolation: for each indicator, non-zero value transformer stations are first selected to form a ranking pool for evaluation, and transformer stations with zero values ​​are directly marked with a score of 0 for that indicator; the transformer stations in the ranking pool are arranged in ascending order of indicator value, and the ranking system is used to handle tied values. The score of a single indicator = the current ranking position of the transformer station / the total number of participating transformer stations × 100, and the score range is 0~100 points. The more serious the problem, the higher the score.

5. The data-driven rural power distribution network substation renovation priority evaluation system according to claim 1, characterized in that, The differentiated weighting system for the weighting configuration and comprehensive score calculation module is as follows: bare conductor ratio 25%, voltage quality 25%, power supply reliability 20%, transformer load rate 15%, and equipment service life 15%. The comprehensive evaluation score is the sum of the normalized scores of the five indicators and their corresponding weights. The higher the score, the higher the priority of the transformation. The system supports the storage of multiple weighting schemes and one-click switching to adapt to different special transformation evaluation scenarios.

6. The data-driven rural power distribution network area renovation priority evaluation system according to claim 1, characterized in that, The priority sorting output and dynamic management module adopts a multi-level sorting rule: the core sorting basis is the comprehensive score in descending order. When the comprehensive scores are the same, the secondary sorting is carried out in turn according to the bare conductor ratio score, voltage quality score, and power supply reliability score. It supports multi-dimensional filtering and sorting by power supply station, administrative region, and problem type, and can generate a special transformation priority list for the corresponding dimension.

7. The data-driven rural power distribution network area renovation priority evaluation system according to claim 1, characterized in that, The priority sorting output and dynamic management module has a dual dynamic update mechanism of regular iteration and event-triggered iteration: it automatically pulls the latest 12 months of operating data at a fixed annual cycle to complete the annual reassessment of all transformer areas and update the annual renovation priority list; when a transformer area completes renovation acceptance, experiences a major fault, or the ledger information is updated, it supports triggering batch real-time reassessment of a single transformer area or a designated area and adjusting the priority ranking in real time.

8. The data-driven rural power distribution network substation renovation priority evaluation system according to claim 1, characterized in that, The priority sorting output and dynamic management module supports dual-format display: list details and GIS spatial heat map. The list displays basic information, original indicators, individual scores, comprehensive scores and problem tags of the transformer area; the GIS map marks the priority level of the transformer area with different colors, intuitively presenting the spatial distribution of high-risk transformer areas. It supports exporting standardized Excel reports containing complete evaluation fields, directly supporting the renovation project initiation and funding allocation.

9. The data-driven rural power distribution network substation renovation priority evaluation system according to claim 1, characterized in that, It adopts a modular loosely coupled architecture, with reserved interfaces for index expansion, external system integration, and algorithm upgrades. It supports the addition of assessment indicators such as three-phase imbalance and low-voltage line loss rate, and can be integrated with project management systems and operation and maintenance control systems to achieve management of assessment, project initiation, implementation, and post-assessment. It is also compatible with the future business expansion needs of rural power distribution network management.

10. A data-driven method for prioritizing the renovation of rural power distribution transformer substations, as described in any one of claims 1-9, characterized in that... include: Step 1 involves multi-source data access and preprocessing, automatically connecting to three heterogeneous data sources: GIS topology, electricity consumption data collection, and PMS equipment ledger. After multi-level cleaning and treatment, the data is deduplicated, outliers are removed, missing values ​​are filled, and cross-source logic is verified. The data is aggregated according to the unique identifier of the transformer area to form a standardized evaluation dataset. A dual-layer storage architecture is adopted to balance full data traceability and batch calculation efficiency. Step two involves calculating core indicators and determining thresholds. Five rural power distribution network-specific evaluation indicators are calculated in batches based on a standardized dataset. Each indicator is then judged against the preset thresholds. Indicators without corresponding hidden dangers are directly assigned a value of 0, while indicators with problems retain their original quantitative values ​​and proceed to the next step. Step 3 is the indicator normalization process, which involves isolating individual indicators at zero values ​​and arranging non-zero indicators in ascending order of value. The relative ranking is linearly mapped to a standard score of 0 to 100, eliminating the difference in units and achieving fair comparison across different regions. Step four involves weight configuration and comprehensive score calculation. Differentiated weights are configured in accordance with the principle of prioritizing safety and people's livelihood. The scores of the five indicators are normalized and weighted to obtain the comprehensive evaluation score. Multiple weight schemes can be switched to adapt to different scenarios. Step five involves priority sorting and dynamic management. A priority list is generated in descending order of comprehensive score, supporting multiple display formats and report export. It also has dynamic iteration capabilities for both periodic and real-time re-evaluation and retains historical archives to support effect traceability.

Images