摘要
针对气体绝缘开关设备(gas insulated switchgear,简称GIS)断路器声信号易受强背景噪声干扰且特征提取困难的问题,提出一种可以有效诊断其机械故障的方法。首先,使用多通道声传感器阵列采集断路器原始观测信号,利用独立成分分析(independent component analysis,简称ICA)方法将观测信号分离为多维源信号,并选取源信号中模糊熵最小的分量作为特征信号;其次,计算特征信号的多尺度模糊熵(multi‑scale fuzzy entropy,简称MFE)生成断路器的声纹特征;最后,利用极限学习机(extreme learning machine,简称ELM)算法识别断路器的故障。实验结果表明,基于声信号的检测方法为GIS断路器的机械故障诊断提供了一种新的解决方案,所提出的算法能够有效提取声纹特征,故障诊断准确率较传统方法有明显提高。
An effective method to diagnose the mechanical faults of gas insulated switchgear(GIS)circuit breakers is proposed to solve the problem,that the sound signals are easily interfered by strong background noise,and the feature extraction is difficult.Firstly,the original observation signals are collected by a multichannel acoustic sensor array,the observation signals are separated into multi-dimensional source signals by independent component analysis(ICA),and the least fuzzy entropy component of the source signal is selected as the characteristic signal.Secondly,the multi-scale fuzzy entropy(MFE)of the characteristic signal is calculated to generate the voiceprint feature of the circuit breaker.Lastly,the extreme learning machine(ELM)algorithm is used to identify the fault.The experimental results show that the detection method based on acoustic signals provides a new solution for GIS circuit breaker mechanical fault diagnosis.The proposed algorithm can effec‑tively extract voiceprint features,and the fault diagnosis accuracy is significantly improved compared with tradi-tional methods.
作者
李可
姚忠远
王枭
顾杰斐
宿磊
薛志钢
LI Ke;YAO Zhongyuan;WANG Xiao;GU Jiefei;SU Lei;XUE Zhigang(Jiangsu Key Laboratory of Advanced Food Manufacturing Equipment and Technology,Jiangnan University Wuxi,214122,China;Shanghai Rhythm Electronic Technology Co.,Ltd.Shanghai,201108,China;Jiangsu Province Special Equipment Safety Supervision Inspection Institute Branch of Wuxi Wuxi,214071,China)
出处
《振动．测试与诊断》
EI
CSCD
北大核心
2023年第5期866-871,1034,1035,共8页
Journal of Vibration,Measurement & Diagnosis
基金
国家自然科学基金资助项目(51705203,51775243,11902124)
江苏省市场监督管理局科技计划资助项目(KJ196043)。
关键词
气体绝缘开关设备断路器
机械故障诊断
声纹
独立成分分析
多尺度模糊熵
极限学习机
gas insulated switchgear(GIS)circuit breaker
mechanical fault diagnosis
voiceprint
independent component analysis
multi-scale fuzzy entropy
extreme learning machine
