Search

CN-122017432-A - Automatic low-voltage feeder cabinet testing system and method based on artificial intelligence

CN122017432ACN 122017432 ACN122017432 ACN 122017432ACN-122017432-A

Abstract

The invention discloses an artificial intelligence-based automatic test system and method for a low-voltage feeder cabinet, which relate to the technical field of intelligent detection of electric power systems and comprise the steps of extracting equipment parameters in an initial knowledge graph by utilizing a neuromorphic calculation engine and dynamically generating an antagonism test waveform by driving a pulse neural network; injecting the antagonism test waveform into a corresponding drawer output terminal of the cabinet to be tested through a dynamic load simulator, and synchronously collecting response signals of current waveform, temperature distribution and action time sequence of the circuit breaker; setting a fault judgment threshold according to the historical sample and the real-time running state, generating a fault diagnosis sub-report and a health state based on the comparison result of the fault node score and the fault judgment threshold, and extracting equipment parameters by utilizing a neuromorphic calculation engine and dynamically generating an antagonism test waveform, thereby greatly improving the intelligent level, diagnosis accuracy and running safety of the low-voltage feeder cabinet test.

Inventors

  • CAO YONGJIE
  • WU TING

Assignees

  • 杭州极能数智装备有限公司

Dates

Publication Date
20260512
Application Date
20260316
Priority Date
20250916

Claims (10)

  1. 1. An artificial intelligence-based automatic test method for a low-voltage feeder cabinet is characterized by comprising the following steps of, Scanning a two-dimensional code of a cabinet body to be detected to obtain design drawing information, constructing a cognitive digital twin body and generating an initial knowledge graph; Extracting equipment parameters in the initial knowledge graph by using a neuromorphic calculation engine, and dynamically generating an antagonism test waveform by driving a pulse neural network; injecting the antagonism test waveform into a corresponding drawer output terminal of the cabinet to be tested through a dynamic load simulator, and synchronously collecting response signals of current waveform, temperature distribution and action time sequence of the circuit breaker; Inputting response signals into a cognitive digital twin body to extract space temperature distribution data, activating insulation aging branches of an initial knowledge graph when the local temperature rise rate is detected to exceed a temperature rise threshold value, extracting temperature rise abnormality indexes, current disturbance indexes and action time difference indexes, and calculating fault node scores; setting a fault judgment threshold according to the historical sample and the real-time running state, and generating a fault diagnosis sub-report and a health state based on a comparison result of the fault node score and the fault judgment threshold; And integrating the fault diagnosis sub-reports and the health states of all drawers, calculating the health index and the overall health index of each drawer, and generating a test report.
  2. 2. The automatic test method of the low-voltage feeder cabinet based on artificial intelligence of claim 1, wherein the method comprises the steps of scanning two-dimensional codes of a cabinet body to be tested to obtain design drawing information, constructing a cognitive digital twin body and generating an initial knowledge graph, Scanning a two-dimensional code of a cabinet body to be detected to obtain design drawing information to obtain equipment parameters, driving a neural symbol combination engine based on the equipment parameters, extracting drawing visual features through a visual neural network, reasoning equipment types and connection relations by combining first-order logic rules, and generating a preliminary knowledge graph; inputting the equipment topological relation in the preliminary knowledge graph into a quantum annealing optimizer to obtain equipment connection strength and importance weight to solve an optimal topological structure, and outputting an optimized electrical connection graph; When the equipment parameter conflict is detected, a multi-agent dialect engine is activated, and automatic correction contradiction parameters are obtained by fusing electrical specifications, historical faults and confidence level decisions of physical constraint agents, so that a correction log is generated; And importing the optimized electrical connection map and the correction log into a hypergraph database to construct a cognitive digital twin body, and encoding equipment nodes, corrected parameters and topological relations in the electrical connection map into a structured knowledge network to form an initial knowledge map.
  3. 3. The automatic test method for low-voltage feeder cabinets based on artificial intelligence according to claim 2, wherein the method comprises extracting device parameters from an initial knowledge graph by using a neuromorphic calculation engine, dynamically generating an antagonism test waveform by driving a pulse neural network, Extracting a breaker action time constant and a transformer characteristic frequency from an initial knowledge graph, constructing an equipment characteristic vector, inputting the equipment characteristic vector into a nerve morphology calculation engine, and driving a pulse neural network to generate a fundamental wave test waveform; And injecting true random noise into the fundamental wave test waveform through the quantum random number generator to generate an antagonism test waveform.
  4. 4. The automatic test method for the low-voltage feeder cabinet based on artificial intelligence according to claim 3, wherein the dynamic load simulator injects the antagonism test waveform into the corresponding drawer output terminal of the cabinet to be tested, synchronously collects the response signals of the current waveform, the temperature distribution and the action time sequence of the circuit breaker, specifically comprises the following steps of, Based on the topological parameters of the cabinet body in the cognitive digital twin body, dynamically adjusting the output impedance of the load simulator to be matched with the drawer terminal to be tested; Injecting an antagonism test waveform into the drawer terminal after impedance matching, triggering a three-way sensor array through a quantum synchronous clock, and synchronously collecting a current waveform signal, an infrared thermal imaging temperature distribution signal and an optical fiber sensing circuit breaker action signal; And fusing the current waveform signal, the infrared thermal imaging temperature distribution signal and the optical fiber sensing circuit breaker action signal, and outputting a response signal.
  5. 5. The method for automatically testing the low-voltage feeder cabinet based on the artificial intelligence according to claim 1, wherein the step of calculating the fault node scores comprises the following steps of, Inputting a response signal into a cognitive digital twin body, extracting thermal_map field space temperature distribution data, calculating the temperature rise rate of each coordinate point, and setting a temperature rise threshold according to the thermal collapse critical point of the insulating material; When the temperature rise rate exceeds the temperature rise threshold value, activating an insulation aging branch in the initial knowledge graph, and obtaining a fault node after the insulation aging branch is matched with the node of the initial knowledge graph through heat conduction path tracking; when the dynamic load simulator injects fundamental wave test waveforms, current waveforms and breaker action time sequence data are collected, temperature rise abnormality indexes, current disturbance indexes and action time difference indexes are calculated, and fault node scores are comprehensively obtained.
  6. 6. The method for automatically testing the low-voltage feeder cabinet based on the artificial intelligence of claim 1, wherein the generating the fault diagnosis sub-report and the health state is dynamically setting a fault judgment threshold according to the historical fault occurrence probability of the equipment and the real-time equipment state, generating the fault diagnosis sub-report when the fault node score is larger than the fault judgment threshold, and marking the fault diagnosis sub-report as the health state when the fault node score is smaller than the fault judgment threshold.
  7. 7. The automatic test method for the low-voltage feeder cabinet based on artificial intelligence of claim 6, wherein the fault diagnosis sub-reports and health states of all drawers are integrated, the health index and the overall health index of each drawer are calculated, a test report is generated, and the specific steps are as follows, Extracting fault diagnosis sub-reports and health states of each drawer, and converting the fault node scores into drawer health indexes by adopting a health index calculation formula; and carrying out weighted summarization on the drawer health indexes to obtain an overall health index, classifying health grades, and generating a test report by fusing each drawer fault diagnosis sub report and the health grade.
  8. 8. An artificial intelligence based automatic test system for a low-voltage feeder cabinet is characterized by comprising a knowledge graph module, a waveform module, a signal module, a weight module, a state evaluation module and an integration module, The knowledge graph module is used for scanning the two-dimensional code of the cabinet body to be detected to obtain design drawing information, constructing a cognitive digital twin body and generating an initial knowledge graph; The waveform module is used for extracting equipment parameters in the initial knowledge graph by utilizing the neuromorphic calculation engine and dynamically generating an antagonism test waveform by driving the pulse neural network; the signal module is used for injecting the antagonism test waveform into the corresponding drawer output terminal of the cabinet body to be tested through the dynamic load simulator and synchronously collecting the response signals of the current waveform, the temperature distribution and the action time sequence of the circuit breaker; the weight module is used for inputting the response signals into the cognitive digital twin body to extract space temperature distribution data, activating an insulation aging branch of the initial knowledge graph when the local temperature rise rate is detected to exceed a temperature rise threshold value, extracting a temperature rise abnormality index, a current disturbance index and a motion time difference index, and calculating fault node scores; The state evaluation module is used for setting a fault judgment threshold according to the historical sample and the real-time running state and generating a fault diagnosis sub-report and a health state based on a comparison result of the fault node score and the fault judgment threshold; The integration module is used for integrating the fault diagnosis sub-reports and the health states of all drawers, calculating the health index and the whole health index of each drawer and generating a test report.
  9. 9. A computer device comprises a memory and a processor, wherein the memory stores a computer program, and the computer program is characterized in that the processor realizes the steps of the artificial intelligence-based automatic testing method for the low-voltage feeder cabinet according to any one of claims 1-7 when executing the computer program.
  10. 10. A computer readable storage medium, on which a computer program is stored, is characterized in that the computer program, when being executed by a processor, implements the steps of the artificial intelligence based automatic test method for low voltage feeder cabinets according to any one of claims 1 to 7.

Description

Automatic low-voltage feeder cabinet testing system and method based on artificial intelligence Technical Field The invention relates to the technical field of intelligent detection of power systems, in particular to an automatic test system and method for a low-voltage feeder cabinet based on artificial intelligence. Background In the process of intelligent transformation of the power system, the low-voltage feeder cabinet is used as a key terminal control and protection device in the power distribution network, and the running reliability of the low-voltage feeder cabinet directly influences the stability of the whole system. With the development of industrial 4.0 and digital twin technology, the traditional low-voltage feeder cabinet testing method which relies on manual detection and static testing means is gradually replaced by an automatic and intelligent scheme. In recent years, intelligent diagnosis technology based on data driving has advanced in the field, such as collecting electrical parameters in real time by adopting a sensor network, classifying and identifying the state of a circuit breaker by combining a machine learning algorithm, and partial research introduces a preliminary digital twin model for simulating equipment behaviors and predicting potential faults. Although the prior art improves the testing efficiency and accuracy of the low-voltage feeder cabinet to a certain extent, the method has obvious limitation in the aspect of dealing with nonlinear response and early fault identification under complex working conditions, the current main stream method is usually excited by depending on preset testing waveforms, and is difficult to cover extreme or antagonistic working condition scenes possibly occurring in actual operation, so that the testing coverage rate is insufficient, most systems only pay attention to single parameter threshold judgment, and lack fusion analysis mechanisms on multi-source heterogeneous signals (such as current waveforms, temperature distribution and time sequence actions), so that the sensitivity to progressive faults such as insulation aging and the like is limited. Disclosure of Invention The present invention has been made in view of the above-described problems occurring in the prior art. Therefore, the invention provides an artificial intelligence-based automatic testing method for the low-voltage feeder cabinet, which solves the problems of insufficient testing coverage rate and poor adaptability to complex working conditions. In order to solve the technical problems, the invention provides the following technical scheme: The invention provides an automatic test method of a low-voltage feeder cabinet based on artificial intelligence, which comprises the steps of scanning a two-dimensional code of a cabinet body to be tested to obtain design drawing information, constructing a cognitive digital twin body and generating an initial knowledge graph, utilizing a neuromorphic calculation engine to extract equipment parameters in the initial knowledge graph and dynamically generating an antagonism test waveform through a driving pulse neural network, injecting the antagonism test waveform into a corresponding drawer output terminal of the cabinet body to be tested through a dynamic load simulator, synchronously collecting response signals of current waveforms, temperature distribution and action time sequences of a circuit breaker, inputting the response signals into the cognitive digital twin body to extract space temperature distribution data, activating insulation ageing branches of the initial knowledge graph when the local temperature rise rate is detected to exceed a temperature rise threshold, extracting temperature rise abnormal indexes, current disturbance indexes and action time difference indexes, calculating fault node scores, setting fault judgment thresholds according to historical samples and real-time running states, generating a fault diagnosis sub-report and a health state based on comparison result of the fault node scores and the fault judgment thresholds, integrating the fault diagnosis sub-report and the health state of all drawers, calculating each drawer health index and the whole health index, and generating a test report. The invention relates to an artificial intelligence based low-voltage feeder cabinet automatic test method, which comprises the steps of scanning two-dimension codes of a cabinet body to be tested to obtain design drawing information, constructing a cognitive digital twin body and generating an initial knowledge graph, Scanning a two-dimensional code of a cabinet body to be detected to obtain design drawing information to obtain equipment parameters, driving a neural symbol combination engine based on the equipment parameters, extracting drawing visual features through a visual neural network, reasoning equipment types and connection relations by combining first-order logic rules, and generating a preliminary knowledg