An adaptive extraction method for faulty power outage lines based on knowledge reasoning

Through knowledge graph and natural language processing technology, the consistency analysis and management of power outage events is carried out, and troubleshooting coefficients and maintenance values ​​are used to optimize troubleshooting and repairs, which solves the problems of inconsistent data and inefficient efficiency of power outage events, and improves troubleshooting efficiency and power supply services. quality.

CN113887823BInactive Publication Date: 2024-11-22赵爱华 +3
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202111226783.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-10-21
Publication Date
2024-11-22
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The existing technology has problems of data inconsistency and inefficiency during the data aggregation and troubleshooting of power outage events, resulting in long power outages and low maintenance work efficiency, affecting the quality of power supply services.

Method used

Adaptive extraction method of fault power outage lines based on knowledge graph and natural language processing is adopted, and the power outage event records of multiple systems are analyzed and managed consistently. The troubleshooting coefficient and maintenance sequence value are used to optimize troubleshooting and maintenance task arrangements to achieve rapid fault finding Line and proactively prevent.

Benefits of technology

It improves data governance efficiency and troubleshooting efficiency, reduces user power outage time, and improves the quality of power supply services.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN113887823B_ABST
    Figure CN113887823B_ABST
Patent Text Reader

Abstract

The invention discloses a method for adaptively extracting faulty power outage lines based on knowledge reasoning, which relates to the technical field of power management, including: based on knowledge graph and natural language processing artificial intelligence technology, realizing consistency analysis and management of power outage event records of multiple systems; when a power outage alarm event occurs, performing troubleshooting coefficient analysis on all power supply lines in the alarm area, and troubleshooting the power supply lines in sequence according to the size of the troubleshooting coefficient; after finding the faulty power outage line, generating a corresponding line maintenance task to a power maintenance center, the power maintenance center performing maintenance sequence value analysis on the received line maintenance task, and reasonably arranging corresponding maintenance personnel to perform maintenance; at the same time, performing power outage linkage analysis on the faulty power outage line, obtaining a corresponding power outage linkage line, and then troubleshooting the power outage linkage line, playing a role of active prevention, further improving fault troubleshooting efficiency, reducing user power outage time, and improving power supply service quality.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of power management, and in particular to a method for adaptively extracting fault power outage lines based on knowledge reasoning. Background Art

[0002] At present, power companies are continuously raising their requirements for data quality and reliability performance indicators of power outage events. In terms of power outage data, due to the differences in business and rules between various production systems and reliability systems, data loss may occur during the aggregation and collection of power outage events, resulting in inconsistencies in the duration of power outages, the scope of power outages, and the number of households at the time of the power outages recorded after the power outage events are collected by various production systems and reliability systems. Manual comparison is required, which is labor-intensive and cumbersome, seriously affecting the efficiency of data governance;

[0003] At the same time, when a power outage alarm occurs, it is difficult to find the fault area and the scope of the power outage. Manual investigation is inefficient, and it is difficult to accurately and quickly obtain the faulty power outage line, which cannot be repaired in time. There is a problem of being unable to reduce the power outage time of users, improve the efficiency of accident repair work, and improve the quality of power supply services. To this end, we propose an adaptive extraction method for faulty power outage lines based on knowledge reasoning. Summary of the invention

[0004] In order to solve the problems existing in the above scheme, the present invention provides a method for adaptive extraction of faulty power outage lines based on knowledge reasoning. The present invention is based on artificial intelligence technologies such as knowledge graphs and natural language processing to achieve consistency analysis and management of power outage event records of multiple systems, ensure data quality, provide a reliable data source for subsequent analysis work, and effectively improve the efficiency of data governance; when a power outage alarm event occurs, all power supply lines in the alarm area are analyzed for troubleshooting coefficients, and the power supply lines are checked in turn according to the size of the troubleshooting coefficients to find the faulty power outage line as soon as possible and improve the troubleshooting efficiency; after finding the faulty power outage line, the faulty power outage line is analyzed for power outage linkage, and then the power outage linkage line is checked, which plays a role of active prevention, further improves the efficiency of fault troubleshooting, reduces the power outage time of users, and improves the quality of power supply services.

[0005] The purpose of the present invention can be achieved through the following technical solutions:

[0006] A method for adaptively extracting faulty power outage lines based on knowledge reasoning includes the following steps:

[0007] Step 1: Based on artificial intelligence technologies such as knowledge graphs and natural language processing, the power outage event records of multiple systems are analyzed and managed in a consistent manner;

[0008] Step 2: When a power outage alarm occurs, perform a troubleshooting coefficient analysis on all power supply lines in the alarm area, and troubleshoot the power supply lines in order according to the troubleshooting coefficients;

[0009] Step 3: After finding the faulty power outage line, generate the corresponding line maintenance task to the power maintenance center, which will perform maintenance sequence value analysis on the received line maintenance task and reasonably arrange the corresponding maintenance personnel to perform maintenance;

[0010] Step 4: After finding the faulty power outage line, perform a power outage linkage analysis on the faulty power outage line to obtain the corresponding power outage linkage line, and then check the power outage linkage line to determine whether a fault problem occurs.

[0011] Furthermore, the specific process of consistency analysis and management of power outage event records of multiple systems is as follows:

[0012] S11: Gathering power outage event records based on text similarity: using natural semantic processing technology to realize the correlation between power outage event records of various systems, integrating power outage event records with text similarity exceeding γ%, and forming a determined power outage event record; wherein γ is a preset value;

[0013] S12: Write power outage event records into the graph database: Use the "entity-relationship" feature in the knowledge graph to store various power outage event records, and support query, maintenance and export through the WEB.

[0014] Furthermore, the graph database is also used to automatically compare the collected power outage event records from several dimensions in combination with the medium-voltage power grid topology to obtain power outage analysis results; and the power outage analysis results are timestamped and stored in the graph database; the dimensions include the duration of the power outage, the scope of the power outage, the number of households at the time of the power outage, and the faulty power outage lines.

[0015] Furthermore, the specific calculation method of the screening coefficient is:

[0016] S21: Obtain all power supply lines in the alarm area; retrieve the power outage analysis results within ninety days before the current system time from the graph database according to the timestamp;

[0017] S22: For the same power supply line, the number of power outages of the power supply line is counted and marked as the power outage frequency P1, the number of power outage durations ≥ the preset duration threshold is counted as the excessive outage frequency P2, the corresponding power outage duration is calculated with the preset duration threshold, and the difference is summed to obtain the excessive outage total value CT; the excessive outage coefficient CS is calculated using the formula CS = P2 × k1 + CT × k2; wherein k1 and k2 are coefficient factors;

[0018] S23: Evaluate the power outage impact value QWi of each power outage according to the power outage impact range and the number of households at the time of the power outage; count the number of times QWi ≥ the impact threshold as the super impact frequency P3, calculate the difference between the corresponding power outage impact value and the impact threshold, and sum the differences to obtain the super impact total value CJ; calculate the super impact coefficient QJ using the formula QJ=P3×k3+CJ×k4, where k3 and k4 are coefficient factors;

[0019] S24: Calculate the screening coefficient PC using the formula PC=P1×g3+CS×g4+QJ×g5, where g3, g4, and g5 are all coefficient factors.

[0020] Furthermore, the specific maintenance process of the maintenance personnel is as follows:

[0021] If the number of line maintenance tasks is one, the maintenance personnel will directly arrive at the corresponding faulty power outage line for maintenance; if the number of line maintenance tasks is greater than one, the line maintenance task with the largest maintenance sequence value will be maintained; after the maintenance is completed, the maintenance sequence values ​​of the remaining line maintenance tasks will be analyzed again, and maintenance will continue to be performed on the line maintenance task with the largest maintenance sequence value, and so on.

[0022] Furthermore, the calculation method of the maintenance sequence value is:

[0023] The distance difference between the line maintenance task location and the maintenance personnel location is obtained as the maintenance distance L1;

[0024] Obtain the faulty power outage line corresponding to the line maintenance task, mark the installation age of the faulty power outage line as N1, mark the total length of the faulty power outage line as L2, and mark the total number of households supplied by the faulty power outage line as ND;

[0025] The maintenance sequence value WX of the corresponding line maintenance task is calculated using the formula WX=(N1×d1+L2×d2+ND×d3) / (L1×d4), where d1, d2, d3, and d4 are all coefficient factors.

[0026] Furthermore, the power outage linkage analysis is carried out on the fault power outage line, specifically:

[0027] Mark the current faulty power outage line as the target line; retrieve the power outage analysis results within 90 days before the current system time from the graph database, and collect the faulty power outage lines in all the power outage analysis results;

[0028] For the same fault power outage line, the number of times the fault power outage line and the target line are synchronously outaged is counted as the synchronous power outage frequency W1; the time difference between the most recent synchronous power outage time of the fault power outage line and the target line and the current system time is calculated to obtain the buffer time HT;

[0029] The power outage linkage value LD of the fault power outage line is calculated using the formula LD=(W1×d5) / (HT×d6), where d5 and d6 are preset coefficient factors; if LD is greater than the linkage threshold, the corresponding fault power outage line is marked as the power outage linkage line of the target line.

[0030] Furthermore, if a certain faulty power outage line and a target line appear in the same power outage analysis result, it is considered that the faulty power outage line and the target line are out of power simultaneously.

[0031] Compared with the prior art, the present invention has the following beneficial effects:

[0032] 1. The present invention is based on artificial intelligence technologies such as knowledge graphs and natural language processing to achieve consistency analysis and management of power outage event records of multiple systems; natural semantic processing technology is used to achieve the correlation between power outage event records of each system, and the power outage event records with text similarity exceeding γ% are integrated to form a confirmed power outage event record; the power outage event record is then written into the graph database; data quality is guaranteed, a reliable data source is provided for subsequent analysis work, and the efficiency of data governance is effectively improved;

[0033] 2. When a power outage alarm occurs, all power supply lines in the alarm area are analyzed for their troubleshooting coefficients, and the power supply lines are checked in order according to the troubleshooting coefficients to find the faulty power outage line as quickly as possible, thus improving the troubleshooting efficiency. The troubleshooting personnel can be general staff members to avoid staff shortages, while reducing the troubleshooting time of maintenance personnel and improving maintenance efficiency.

[0034] 3. After finding the faulty power outage line, the power maintenance center conducts a maintenance sequence value analysis on the received line maintenance tasks, and the maintenance personnel perform maintenance on the line maintenance tasks in sequence according to the size of the maintenance sequence value WX, further improving the maintenance efficiency; at the same time, after finding the faulty power outage line, a power outage linkage analysis is performed on the faulty power outage line to obtain the corresponding power outage linkage line, and then the power outage linkage line is checked to determine whether a fault problem occurs, which plays a role of active prevention, further improving the efficiency of fault investigation, reducing the power outage time for users, and improving the quality of power supply services. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0036] Figure 1 The present invention is a flow chart of the method. DETAILED DESCRIPTION

[0037] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0038] like Figure 1 As shown, a method for adaptively extracting faulty power outage lines based on knowledge reasoning includes the following steps:

[0039] Step 1: Based on artificial intelligence technologies such as knowledge graphs and natural language processing, the power outage event records of multiple systems are analyzed and managed in a consistent manner, including:

[0040] S11: Gathering power outage event records based on text similarity: Gathering power outage event records of multiple systems such as 95598, dispatching, job control, and power consumption information collection, using natural semantic processing technology to realize the correlation between the power outage event records of the power consumption information collection system and the power outage event records of other systems, integrating the power outage event records with text similarity exceeding γ%, and forming a confirmed power outage event record; wherein γ is a preset value;

[0041] S12: Write power outage event records into the graph database: Use the "entity-relationship" feature in the knowledge graph to store various power outage event records, and support query, maintenance and export through the WEB;

[0042] Combined with the medium-voltage power grid topology, the collected power outage event records are automatically compared from several dimensions, such as power outage duration, power outage impact range, number of households at the time of the power outage, and faulty power outage lines, to obtain the power outage analysis results. The power outage analysis results are timestamped and stored in the graph database.

[0043] The power outage analysis results include the features of each power outage dimension, which are presented in visual forms such as graphics and charts, and interactively processed to enhance the data presentation effect, so that users can more conveniently and intuitively observe and understand the data, and then discover the hidden information in the data; the power outage dimension features include the power outage duration, the scope of the power outage, the number of households at the time of the power outage, the faulty power outage line, etc.; one power outage event can contain multiple faulty power outage lines; and multiple faulty power outage lines can be distributed in different areas;

[0044] Based on artificial intelligence technologies such as knowledge graphs and natural language processing, the present invention realizes consistency analysis and management of power outage event records of multiple systems, ensures data quality, provides a reliable data source for subsequent analysis work, and effectively improves the efficiency of data management;

[0045] Step 2: When a power outage alarm occurs, perform a troubleshooting coefficient analysis on all power supply lines in the alarm area, and check the power supply lines in order according to the troubleshooting coefficients to find the faulty power outage line. The specific steps are as follows:

[0046] S21: Obtain all power supply lines in the alarm area; retrieve the power outage analysis results within ninety days before the current system time from the graph database according to the timestamp;

[0047] S22: for the same power supply line, according to the power outage analysis result, the number of power outages of the power supply line is counted and marked as the power outage frequency P1, and the power outage duration of each power outage of the power supply line is compared with the preset duration threshold; if the power outage duration ≥ the preset duration threshold, the corresponding power outage duration is marked as excessive power outage duration;

[0048] The number of occurrences of the over-stop duration is counted as the over-stop frequency P2, the difference between the over-stop duration and the preset duration threshold is calculated, and the corresponding differences are summed to obtain the total over-stop value CT;

[0049] The frequency of over-stop and the total value of over-stop are normalized and their values ​​are taken, and the over-stop coefficient CS is calculated using the formula CS=P2×k1+CT×k2; where k1 and k2 are both preset coefficient factors;

[0050] S23: The power outage impact range of each power outage of the power supply line is marked as Wi, and the number of households during the power outage is marked as Hi; the power outage impact range is represented by the size of the power outage area (in square meters); the number of households during the power outage is represented by the number of households in the power outage area; i represents the i-th power outage;

[0051] The power outage impact value QWi of each power outage is evaluated according to the power outage impact range and the number of households at the time of the power outage, specifically:

[0052]

[0053] In the formula, g1 and g2 are preset coefficient factors;

[0054] Compare the power outage impact value QWi with the impact threshold; count the number of times QWi ≥ the impact threshold as the excess impact frequency P3, calculate the difference between the corresponding power outage impact value and the impact threshold, and sum the differences to obtain the excess impact total value CJ;

[0055] The frequency of super-contact and the total value of super-contact are normalized and their values ​​are taken, and the super-contact coefficient QJ is calculated using the formula QJ=P3×k3+CJ×k4, where k3 and k4 are coefficient factors;

[0056] S24: normalize the power outage frequency, the over-outage coefficient, and the over-contagion coefficient and take their values, and calculate the troubleshooting coefficient PC using the formula PC=P1×g3+CS×g4+QJ×g5, where g3, g4, and g5 are all coefficient factors;

[0057] Since checking the power supply line and judging whether it is a faulty power outage line does not require the checker to have high professional knowledge, the checker can be a general staff member, and then the power supply line is checked in turn according to the size of the check coefficient PC, so as to find the faulty power outage line as soon as possible, thereby improving the checking efficiency; after finding the faulty power outage line, it will be repaired by professional circuit maintenance personnel, avoiding the shortage of maintenance personnel, and at the same time reducing the fault checking time of circuit maintenance personnel, thereby improving the maintenance efficiency;

[0058] Step 3: After finding the faulty power outage line, generate a corresponding line maintenance task to the power maintenance center. The line maintenance task carries the location information of the faulty power outage line.

[0059] The power maintenance center analyzes the received line maintenance tasks and reasonably arranges the corresponding maintenance personnel to carry out maintenance; the specific maintenance process is as follows:

[0060] If the number of line maintenance tasks is one, the maintenance personnel will directly go to the corresponding faulty power outage line to perform maintenance;

[0061] If the number of line maintenance tasks is greater than one, a maintenance sequence value analysis is performed on several line maintenance tasks, and the line maintenance task with the largest maintenance sequence value is first maintained. After the maintenance is completed, the maintenance sequence value analysis is performed again on the remaining line maintenance tasks, and the line maintenance task with the largest maintenance sequence value is continued to be maintained, and so on. In the present invention, the maintenance personnel repair the line maintenance tasks in sequence according to the size of the maintenance sequence value WX, so as to further improve the maintenance efficiency;

[0062] The calculation method of maintenance sequence value is:

[0063] Calculate the distance difference between the position of the line maintenance task and the position of the maintenance personnel to obtain the maintenance distance L1;

[0064] Obtain the faulty power outage line corresponding to the line maintenance task, mark the installation age of the faulty power outage line as N1, mark the total length of the faulty power outage line as L2, and mark the total number of households supplied by the faulty power outage line as ND;

[0065] The maintenance sequence value WX of the corresponding line maintenance task is calculated using the formula WX = (N1 × d1 + L2 × d2 + ND × d3) / (L1 × d4), where d1, d2, d3, and d4 are all coefficient factors;

[0066] In this embodiment, the method further includes: after finding the faulty power outage line, performing a power outage linkage analysis on the faulty power outage line to obtain a corresponding power outage linkage line, and then troubleshooting the power outage linkage line to determine whether a fault problem occurs, thereby playing a proactive preventive role and further improving the efficiency of troubleshooting; specifically:

[0067] Retrieve the power outage analysis results within 90 days before the current system time from the graph database;

[0068] Mark the currently found faulty power outage line as the target line, collect the faulty power outage lines in all power outage analysis results, and if a faulty power outage line and the target line appear in the same power outage analysis result, it is considered that the faulty power outage line and the target line are out of power simultaneously;

[0069] For the same fault power outage line, the number of times the fault power outage line and the target line are synchronously outaged is counted as the synchronous power outage frequency W1; the time difference between the most recent synchronous power outage time of the fault power outage line and the target line and the current system time is calculated to obtain the buffer time HT;

[0070] The power outage linkage value LD of the fault power outage line is calculated using the formula LD = (W1 × d5) / (HT × d6), where d5 and d6 are preset coefficient factors;

[0071] The power outage linkage value LD is compared with the linkage threshold. If LD is greater than the linkage threshold, the corresponding fault power outage line is marked as the power outage linkage line of the target line.

[0072] The above formulas are all calculated by removing dimensions and taking numerical values. The formula is a formula that is closest to the actual situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained by simulating a large amount of data.

[0073] Working principle of the present invention:

[0074] A method for adaptive extraction of faulty power outage lines based on knowledge reasoning. When working, the power outage event records are first aggregated based on text similarity, and the power outage event records of multiple systems such as 95598, dispatching, operation control, and electricity consumption information collection are aggregated together. The power outage event records with text similarity exceeding γ% are integrated to form a confirmed power outage event record; then the power outage event record is written into the graph database, and the "entity-relationship" feature in the knowledge graph is used to store various power outage event records. Combined with the medium-voltage power grid topology, the aggregated power outage event records are automatically compared from several dimensions such as power outage duration, power outage impact range, number of households during power outage, and faulty power outage line to obtain the power outage analysis result; and the power outage analysis result is timestamped and stored in the graph database to ensure data quality, provide a reliable data source for subsequent analysis work, and effectively improve the efficiency of data governance;

[0075] When a power outage alarm occurs, a troubleshooting coefficient analysis is performed on all power supply lines in the alarm area, and the power supply lines are checked in turn according to the size of the troubleshooting coefficient to find the faulty power outage line, thereby improving the troubleshooting efficiency; after finding the faulty power outage line, a corresponding line maintenance task is generated and sent to the power maintenance center, which performs a maintenance sequence value analysis on the received line maintenance tasks, and the maintenance personnel repair the line maintenance tasks in turn according to the size of the maintenance sequence value WX, thereby further improving the maintenance efficiency; at the same time, after finding the faulty power outage line, a power outage linkage analysis is performed on the faulty power outage line to obtain the corresponding power outage linkage line, and then the power outage linkage line is checked to determine whether a fault problem occurs, thereby playing a role of active prevention, further improving the efficiency of fault troubleshooting, reducing the power outage time for users, and improving the quality of power supply services.

[0076] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0077] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for adaptively extracting faulty power outage lines based on knowledge reasoning, characterized in that: The steps include: Step 1: Based on knowledge graph and natural language processing technology, realize consistency analysis and management of power outage event records of multiple systems; Step 2: When a power outage alarm occurs, perform a troubleshooting coefficient analysis on all power supply lines in the alarm area, and troubleshoot the power supply lines in turn according to the troubleshooting coefficients. The specific steps are as follows: S21: Obtain all power supply lines in the alarm area; According to the timestamp, retrieve the power outage analysis results within 90 days before the current system time from the graph database; S22: for the same power supply line, according to the power outage analysis result, the number of power outages of the power supply line is counted and marked as the power outage frequency P1, and the power outage duration of each power outage of the power supply line is compared with the preset duration threshold; if the power outage duration ≥ the preset duration threshold, the corresponding power outage duration is marked as excessive power outage duration; The number of occurrences of the over-stop duration is counted as the over-stop frequency P2, the difference between the over-stop duration and the preset duration threshold is calculated, and the corresponding differences are summed to obtain the total over-stop value CT; The frequency of over-stop and the total value of over-stop are normalized and their values ​​are taken, and the over-stop coefficient CS is calculated using the formula CS=P2×k1+CT×k2; Among them, k1 and k2 are preset coefficient factors; S23: Mark the power outage impact range of the power supply line as Wi, and the number of households during the power outage as Hi; The power outage impact range is represented by the size of the power outage area; the number of households during the power outage is represented by the number of households in the power outage area; i represents the i-th power outage; The power outage impact value QWi of each power outage is evaluated according to the power outage impact range and the number of households at the time of the power outage, specifically: In the formula, g1 and g2 are preset coefficient factors; Compare the power outage impact value QWi with the impact threshold; The number of times when QWi ≥ the impact threshold is counted as the excess impact frequency P3, the corresponding power outage impact value and the impact threshold are calculated by difference, and the difference is summed to obtain the excess impact total value CJ; The frequency of super-contact and the total value of super-contact are normalized and their values ​​are taken, and the super-contact coefficient QJ is calculated using the formula QJ=P3×k3+CJ×k4, where k3 and k4 are coefficient factors; S24: normalize the power outage frequency, the over-outage coefficient, and the over-contagion coefficient and take their values, and calculate the troubleshooting coefficient PC using the formula PC=P1×g3+CS×g4+QJ×g5, where g3, g4, and g5 are all coefficient factors; Step 3: After finding the faulty power outage line, generate the corresponding line maintenance task to the power maintenance center. The power maintenance center analyzes the maintenance sequence value of the received line maintenance task and reasonably arranges the corresponding maintenance personnel to perform maintenance. The specific maintenance process of the maintenance personnel is as follows: If the number of line maintenance tasks is one, the maintenance personnel will directly go to the corresponding faulty power outage line for maintenance; if the number of line maintenance tasks is greater than one, the line maintenance task with the largest maintenance sequence value will be repaired; After the maintenance is completed, the maintenance sequence value analysis is performed on the remaining line maintenance tasks again, and the line maintenance task with the largest maintenance sequence value is continued to be repaired, and so on; The calculation method of maintenance sequence value is: Calculate the distance difference between the location of the line maintenance task and the location of the maintenance personnel to obtain the maintenance distance L1; obtain the fault power outage line corresponding to the line maintenance task; The installation age of the faulty power outage line is marked as N1, the total length of the faulty power outage line is marked as L2, and the total number of households supplied by the faulty power outage line is marked as ND; The maintenance sequence value WX of the corresponding line maintenance task is calculated using the formula WX = (N1 × d1 + L2 × d2 + ND × d3) / (L1 × d4), where d1, d2, d3, and d4 are all coefficient factors; Step 4: After finding the faulty power outage line, perform a power outage linkage analysis on the faulty power outage line to obtain the corresponding power outage linkage line, and then check the power outage linkage line to determine whether a fault problem occurs; Specifically: Mark the current fault power outage line as the target line; Retrieve the power outage analysis results within 90 days before the current system time from the graph database, and collect the faulty power outage lines in all the power outage analysis results; For the same fault power outage line, the number of times the fault power outage line and the target line are synchronously powered off is counted as the synchronous power outage frequency W1; If a faulty power outage line and a target line appear in the same power outage analysis result, it is considered that the faulty power outage line and the target line are out of power simultaneously; The buffering time HT is obtained by calculating the time difference between the most recent synchronous power outage time of the fault power outage line and the target line and the current system time; The power outage linkage value LD of the fault power outage line is calculated using the formula LD = (W1 × d5) / (HT × d6), where d5 and d6 are preset coefficient factors; If LD is greater than the linkage threshold, the corresponding fault power outage line will be marked as the power outage linkage line of the target line.

2. According to the method of adaptive extraction of faulty power outage lines based on knowledge reasoning in claim 1, it is characterized in that: The specific process of consistency analysis and management of power outage event records of multiple systems is as follows: S11: Gathering power outage event records based on text similarity: using natural semantic processing technology to realize the correlation between power outage event records of various systems, integrating power outage event records with text similarity exceeding γ%, and forming a determined power outage event record; wherein γ is a preset value; S12: Write power outage event records into the graph database: Use the "entity-relationship" feature in the knowledge graph to store various power outage event records, and support query, maintenance and export through the WEB.

3. The adaptive extraction method of faulty power outage lines based on knowledge reasoning according to claim 2 is characterized in that: The graph database is also used to automatically compare the collected power outage event records from several dimensions in combination with the medium voltage power grid topology to obtain a power outage analysis result; and the power outage analysis result is timestamped and stored in the graph database; The dimensions include the duration of the power outage, the scope of the power outage, the number of households at the time of the power outage, and the faulty power outage lines.