高压电器2025，Vol.61Issue(9)：1-7,17,8.DOI:10.13296/j.1001-1609.hva.2025.09.001
基于小波包能量分解与神经网络的断路器故障诊断研究
Research on Fault Diagnosis of Circuit Breaker Based on Wavelet Packet Energy Decomposition and Neural Network
摘要
Abstract
Circuit breaker,which undertakes critical control and protection functions in power grid,is prone to latent faults due to its complex structure and numerous parts.It is of great significance to strengthen the research on the la-tent fault diagnosis of operating mechanism system for the circuit breaker and find the latent fault in advance so to im-prove the operation reliability of the circuit breaker.In this paper type ZN98 vacuum circuit breaker is taken as the research object and vibration acceleration test system of the vacuum circuit breaker is set up.Moreover,such latent faults as losseness of fastening bolt of buffer,looseness of fastening bolt of hanging plate,change of thimble stroke of opening coil,failure of return spring of closing coil,aging of opening spring,failure of buffer and firmly fastening bolt of closing spring are taken as examples,the fault diagnosis characteristic quantity is extracted based on wavelet pack-et frequency band energy decomposition algorithm and falut identification is performed by the neural network algo-rithm.The research results show that the normalized energy values of different frequency bands after decomposition shall be selected,and the two sensitive frequency bands with the highest discrimination shall be selected as the latent fault diagnosis features,and PNN neural network algorithm has more reliable accuracy and efficiency than those of BP neural network algorithm.Therefore,PNN neural network is more suitable for the practical engineering applica-tion of latent fault identification of high voltage circuti breaker.关键词
高压断路器/潜伏性故障/小波包频带能量分解/PNN神经网络/故障识别Key words
high voltage circuit breaker/latent fault/wavelet packet frequency band energy decomposition/PNN neural network/fault identification引用本文复制引用
鄢呈旸,王立军,张闻哲,张佳灏,林婧..基于小波包能量分解与神经网络的断路器故障诊断研究[J].高压电器,2025,61(9):1-7,17,8.基金项目
国家自然科学基金项目(52377157).Project Supported by National Natural Science Foundation of China(52377157). （52377157）
