摘要
为解决复杂供输弹系统早期微弱特征提取的问题,提出将自适应可调品质因子小波变换(TQWT)和样本熵相结合的供输弹故障诊断方法。基于能量加权归一化小波熵对TQWT的必要参数进行自适应选取,运用样本熵特征指标选出5个最优的特征子带,将样本熵组成的特征向量作为Elman神经网络的输入,对供输弹系统早期故障进行识别,输出结果显示,将TQWT和样本熵相结合的诊断方法可用于供输弹系统故障识别,准确率高达9224%。
Aiming at the problem of early weak feature extraction of complex ammunition supply system,a fault diagnosis method based on adaptive adjustable quality factor wavelet transform(TQWT)and sample entropy is proposedFirst,based on the energy weighted normalized wavelet entropy,the necessary parameters of TQWT are adaptively selectedThen,five optimal feature subbands are selected through the feature index of sample entropy.Finally,the feature vector composed of sample entropy is used as the input of Elman neural network to identify the early fault of ammunition supply systemThe results show that the diagnosis method combining TQWT and sample entropy can be used for fault diagnosis of ammunition supply system with an accuracy rate of 9224%.
作者
赵璐
潘宏侠
许昕
刘燕军
高俊峰
付志敏
ZHAO Lu;PAN Hongxia;XU Xin;LIU Yanjun;GAO Junfeng;FU Zhimin(School of Mechanical Engineering,North University of China,Taiyuan 030051,China;System Identification and Diagnosis Technology Research Institute,North University of China,Taiyuan 030024,China;Inner Mongolia North Heavy Industry Group Research Institute,Baotou Inner Mongolia 014033,China;Inner Mongolia First Machine Group Research Institute,Baotou Inner Mongolia 014032,China;Aviation industry Hongdu 660,Nanchang 330024,China)
出处
《机械设计与研究》
CSCD
北大核心
2021年第1期210-214,共5页
Machine Design And Research
基金
国家自然科学基金项目资助项目(51675491)
面上自然基金项目(201801D121185)
一般项目(JCKY2018408C002)。
关键词
供输弹系统
自适应TQWT
样本熵
故障诊断识别
ammunition supply system
adaptive TQWT
sample entropy
fault diagnosis identification
