摘要
针对多类型故障诊断结果的输出目标值单一、准确度低的问题,提出报修信息生命周期化采集下的电子设备故障快速诊断方法.从电子设备结构维、信息表示维和生命周期维3个方面,构建故障报修信息生命周期模型,依据电子设备运行数据波动程度,确定数据采集间隔,周期化采集电子设备故障信息;采用小波分析技术提取故障特征,通过归一化和降维方法降低故障特征数据维度;设计故障诊断框架,使用模糊神经网络局部诊断电子设备故障,得到n个证据体,运用D-S理论得到故障诊断决策规则,确定电子设备故障类型.实例测试结果表明,此次所提方法准确诊断电子设备中存在的多种故障类型,电路故障输出目标值多样.
Aiming at the problem of single output target value and low accuracy of multi type fault diagnosis results, a fast fault diagnosis method of electronic equipment under the life cycle collection of repair information is proposed. From the three aspects of electronic equipment structure dimension, information representation dimension and life cycle dimension, the life cycle model of fault repair information is constructed, the data collection interval is determined according to the fluctuation degree of electronic equipment operation data, and the fault information of electronic equipment is collected periodically;Wavelet analysis technology is used to extract fault features, and normalization and dimensionality reduction methods are used to reduce the dimension of fault feature data;The fault diagnosis framework is designed. The fuzzy neural network is used to diagnose the fault of electronic equipment locally, and n evidence bodies are obtained. The fault diagnosis decision rules are obtained by using D-S theory to determine the fault type of electronic equipment. The example test results show that the proposed method can accurately diagnose a variety of fault types in electronic equipment, and the output target values of circuit faults are diverse.
作者
杨晓娜
刘志远
王蔚
周佳
许亚伟
邓珂
YANG Xiaona;LIU Zhiyuan;WANG Wei;ZHOU Jia;XU Yawei;DENG Ke(State Grid Qingyang Electric Power Company,Qingyang 745000,China;State Grid Gansu Electric Power Company,Lanzhou 730030,China)
出处
《测试技术学报》
2022年第5期455-460,共6页
Journal of Test and Measurement Technology
关键词
报修信息
生命周期
电子设备
设备故障
故障诊断
故障特征
repair report information
life cycle
electronic equipment
equipment failure
fault diagnosis
fault characteristics
