CN-122020460-A - Distribution box defect identification method, system, equipment and medium

CN122020460ACN 122020460 ACN122020460 ACN 122020460ACN-122020460-A

Abstract

The invention discloses a method, a system and a medium for identifying defects of a distribution box, and belongs to the technical field of distribution box identification, wherein the method comprises the steps of identifying the defects of the distribution box, carrying out association analysis and defect identification on collected environmental parameter data and electrical parameter data through a preset wiring loosening knowledge graph to obtain a first defect identification result, and determining decision feature deviation values; the decision feature deviation value is input into a preset wiring loosening decision tree to conduct defect identification to obtain a second defect identification result, and comprehensive analysis is conducted according to the first defect identification result and the second defect identification result to obtain a final defect identification result.

Inventors

LIU GAO
GUO JINCHAO
LIN JUNSHENG
LUO JINBIN
CHEN HAO
YANG KE
XU JIE
LI QIYAO

Assignees

广东电网有限责任公司

Dates

Publication Date: 20260512
Application Date: 20260122

Claims (10)

1. A method for identifying defects of a distribution box, comprising: performing association analysis and defect identification on the collected environmental parameter data and electrical parameter data through a preset wiring loosening knowledge graph to obtain a first defect identification result; Determining a decision feature deviation value according to the electrical parameter data and combining a parameter reference value in the wire loosening knowledge graph, inputting the decision feature deviation value into a preset wire loosening decision tree for defect identification, and obtaining a second defect identification result; And comprehensively analyzing according to the first defect identification result and the second defect identification result to obtain a final defect identification result.
2. The method for identifying defects of a distribution box according to claim 1, wherein the correlation analysis and the defect identification are performed on the collected environmental parameter data and the collected electrical parameter data through a preset wiring loosening knowledge graph to obtain a first defect identification result, specifically: Performing preliminary association matching on a parameter characteristic range in a preset wiring loosening knowledge graph and acquired environment parameter data and electrical parameter data to obtain a preliminary matching wiring state group corresponding to the environment parameter data and the electrical parameter data respectively, wherein the electrical parameter data comprises terminal temperature data, terminal current data and terminal vibration frequency data; According to the preliminary matching wiring state group, counting the parameter coincidence item number of the environmental parameter data and the electrical parameter data, and performing comparison analysis to obtain first matching credibility data; Screening the matched reliability data through a preset reliability threshold value to obtain a parameter wiring state matching pair, constructing a first collaborative matching relationship through the wiring loosening knowledge graph, and carrying out consistency comparison on the first collaborative matching relationship to obtain a first consistency result; and performing defect identification based on the first consistency result, the first cooperative matching relation, the first matching reliability data and the wiring looseness knowledge graph to obtain a first defect identification result.
3. The method for identifying defects of a distribution box according to claim 2, wherein the defect identification is performed based on the first consistency result, the first collaborative matching relationship, the first matching reliability data and the wire loosening knowledge graph to obtain a first defect identification result, specifically: Screening the parameter wiring state matching pairs based on the first consistency result to obtain wiring state candidate group data, wherein the wiring state candidate group data comprises corresponding collaborative matching relations and corresponding matching credibility data corresponding to each candidate wiring state; Determining the association influence degree of each candidate wiring state by other parameters and other wiring states through the wiring loosening knowledge graph based on the wiring state candidate group data; Based on the association influence degree and the first matching credibility data, carrying out priority ranking on the candidate wiring states in the wiring state candidate group data to obtain a candidate wiring state list; and selecting a candidate wiring state with the highest priority based on the candidate wiring state list, and carrying out defect identification through the wiring loosening knowledge graph to obtain a first defect identification result of the distribution box.
4. The method for identifying defects of a distribution box according to claim 1, wherein the step of inputting the decision feature deviation value into a preset wire loosening decision tree to identify defects, and obtaining a second defect identification result comprises the following steps: Classifying the decision feature deviation values according to the electrical parameter data, and removing abnormal values to obtain an effective decision feature deviation value group; mapping the effective decision feature deviation value group through a preset wiring loosening decision tree, constructing a deviation interval identification combination, carrying out relevance judgment on the deviation interval identification combination through the wiring loosening knowledge graph, and determining a terminal to be subjected to important verification; And acquiring a deviation change rate sequence of each terminal to be subjected to important verification in a preset time period according to the terminal to be subjected to important verification, and analyzing through the wire loosening decision tree according to the deviation change rate sequence to acquire a second defect identification result.
5. The method for identifying defects of a distribution box according to claim 4, wherein the analysis is performed through the wire loosening decision tree according to the deviation change rate sequence to obtain a second defect identification result, specifically: Comparing and analyzing the deviation change rate sequence through the wiring loosening decision tree, and fusing according to an anomaly fusion rule to obtain comprehensive deviation change anomaly data of the terminal to be subjected to important verification; carrying out data analysis on the comprehensive deviation change anomaly data through the wiring loosening decision tree, and counting the anomaly class identification of the terminal to be subjected to key check to obtain an anomaly class terminal distribution statistical result; Comparing and analyzing the abnormal grade identification with the abnormal grade terminal distribution statistical result through the wiring loosening decision tree to obtain first defect tendency type data of the terminal to be checked, and correcting the first defect tendency type data through the wiring loosening knowledge graph to obtain second defect tendency type data; And carrying out defect confirmation on the key checking terminal through the wiring loosening decision tree according to the second defect tendency type data, and integrating a normal state judgment result of a normal terminal to obtain a second defect identification result.
6. The method for identifying defects of a distribution box according to claim 1, wherein the first defect identification result comprises a wire looseness judging level, associated analysis electrical parameter data, matching degree with the wire looseness knowledge graph and defect influence degree data of environmental parameter data; The second defect identification result comprises a wiring loosening judgment level, a decision feature deviation value, a deviation degree of a parameter reference value in the wiring loosening knowledge graph and sensitivity of the wiring loosening decision tree to defect identification; The step of comprehensively analyzing the first defect recognition result and the second defect recognition result to obtain a final defect recognition result comprises the following steps: Constructing a first result credibility value according to the first defect identification result, constructing a second result credibility value according to the second defect identification result, and combining a result association coefficient according to the first result credibility value and the second result credibility value to obtain comprehensive association credibility; And carrying out consistency analysis on the first defect identification result and the second defect identification result according to the comprehensive correlation credibility to obtain a final defect identification result.
7. The method for identifying defects of a distribution box according to claim 6, wherein the consistency analysis is performed on the first defect identification result and the second defect identification result according to the comprehensive correlation reliability, so as to obtain a final defect identification result, specifically: determining a first consistency coefficient of a first defect identification result based on the types of the wiring loosening defects corresponding to different electrical parameter data in the first defect identification result; Determining a second consistency coefficient of a second defect identification result based on the wiring loosening defect subtype corresponding to different decision feature deviation values in the second defect identification result; analyzing based on the first consistency coefficient, the second consistency coefficient and the comprehensive association credibility to determine the comprehensive consistency credibility of the first defect identification result and the second defect identification result; And according to the comprehensive consistency reliability, comparing and analyzing through a preset defect identification result judgment threshold value to obtain a final defect identification result.
8. The distribution box defect identification system is characterized by comprising a first identification module, a second identification module and a comprehensive analysis module; The first identification module is used for carrying out association analysis and defect identification on the collected environmental parameter data and electrical parameter data through a preset wiring loosening knowledge graph to obtain a first defect identification result; the second recognition module is used for determining a decision feature deviation value according to the electrical parameter data and combining a parameter reference value in the wire loosening knowledge graph, inputting the decision feature deviation value into a preset wire loosening decision tree for defect recognition, and obtaining a second defect recognition result; And the comprehensive analysis module is used for comprehensively analyzing according to the first defect identification result and the second defect identification result to obtain a final defect identification result.
9. A terminal device comprising a processor, a memory and a computer program stored in the memory and configured to be executed by the processor, the processor implementing the method of identifying a defect of an electrical distribution box according to any of claims 1-7 when the computer program is executed by the processor.
10. A computer readable storage medium comprising a stored computer program, wherein the computer program, when run, controls a device in which the computer readable storage medium is located to perform the method of identifying a defect of an electrical distribution box as claimed in any one of claims 1-7.

Description

Distribution box defect identification method, system, equipment and medium Technical Field The invention relates to the technical field of distribution box identification, in particular to a distribution box defect identification method, a distribution box defect identification system, a distribution box defect identification device and a distribution box defect identification medium. Background In the daily operation and maintenance of a power system, the distribution box is used as key equipment for power distribution and control, defects caused by loosening of internal wiring of the distribution box can directly lead to increase of contact resistance and abnormal local heating, and safety accidents such as arc discharge, equipment burnout and even electric fire can be possibly caused when the defects are serious, so that the accurate identification of the loosening defects of the distribution box wiring is very important. The existing distribution box defect identification method mostly adopts a mode of combining manual inspection and sensor monitoring, wherein the manual inspection judges whether loose defects exist or not by operation and maintenance personnel according to visual observation of the color change of a wiring terminal, manual touch sensing of temperature difference or probing of wiring fastening degree by a screwdriver, the mode is seriously dependent on personnel experience and has strong subjectivity, the problems of missed judgment, misjudgment and low inspection efficiency are easy to occur, when the sensor monitors, only a single physical signal is generally adopted, and whether the wiring is loosened is judged based on a preset threshold value, so that the single physical signal is not only easily interfered by environmental factors, and the misjudgment rate is high, but also the multidimensional information related to the loosening of the wiring cannot be correlated, the loosening state of the wiring is difficult to comprehensively reflect, the accuracy of defect identification is insufficient, and the identification of the loosening defect of the wiring of the distribution box cannot be efficiently and accurately completed. Disclosure of Invention The invention provides a method, a system, equipment and a medium for identifying defects of a distribution box, which can analyze the wiring state of the distribution box by combining a wiring loosening knowledge graph so as to realize high-precision defect identification of the distribution box. The invention provides a defect identification method of a distribution box, which comprises the following steps: performing association analysis and defect identification on the collected environmental parameter data and electrical parameter data through a preset wiring loosening knowledge graph to obtain a first defect identification result; Determining a decision feature deviation value according to the electrical parameter data and combining a parameter reference value in the wire loosening knowledge graph, inputting the decision feature deviation value into a preset wire loosening decision tree for defect identification, and obtaining a second defect identification result; And comprehensively analyzing according to the first defect identification result and the second defect identification result to obtain a final defect identification result. According to the embodiment of the invention, the environmental parameter data and the electrical parameter data are acquired, the limitation of single parameter analysis is broken through by combining the preset wire looseness knowledge graph to obtain the first defect identification result fused with multi-dimensional information, the information correlation mining capability and defect identification accuracy are enhanced, the decision characteristic deviation value is determined according to the electrical parameter data by combining the parameter reference value in the wire looseness knowledge graph and is input into the preset wire looseness decision tree, the second defect identification result containing the parameter deviation logic is obtained by hierarchically judging the characteristic deviation by means of the decision tree, the defect of a single judging method is overcome by comprehensively analyzing the first defect identification result and the second defect identification result and combining the double advantages of correlation analysis and deviation logic judgment, the defect identification result containing the wire looseness information of the distribution box is finally output, the problem of missed judgment and misjudgment caused by strong experience and subjectivity of manual inspection is effectively solved, the inspection efficiency is improved, and meanwhile, the defect that the single physical signal is easy to be interfered by environment is avoided by multi-source data correlation and multi-level judgment, and the high efficiency of inspection is realized. Further, thr