摘要
针对齿轮箱故障信号的复杂性和非平稳性,提出了一种基于小波分解和样本熵的遗传算法支持向量机(GA-SVM)故障诊断方法。采用小波分解对信号进行三层分解并提取其高频系数与低频系数,然后计算其系数的样本熵值并构建特征向量,最后将其输入到经过遗传算法(GA)优化后的支持向量机中进行识别。实验表明,对4种工况下6类齿轮箱状态样本进行分类,通过GA算法优化后的SVM模型具有较高的识别准确率且高于文中其他识别模型。
For the complexity and non-stationarity of gearbox fault signals,a method of genetic algorithm support vector machine(GA-SVM)based on wavelet decomposition and sample entropy was proposed.Using wavelet decomposition to decompose the signal into three layers and extracting its high frequency coefficient and low frequency coefficient.Then the sample entropy of its coefficient is calculated and the eigenvector is constructed.Finally,it is input into the support vector machine optimized by genetic algorithm(GA).Experimental results show that the six types of gearbox state samples under four conditions are classified and the SVM model optimized by GA algorithm has higher recognition accuracy than other recognition models in the paper.
作者
姜保军
曹浩
JIANG Bao-jun;CAO Hao(School of Mechanotronics&Vehicle Engineering,Chongqing Jiaotong University,Chongqing 400074,China;Chongqing Key Laboratory of System Integration and Control for Urban Rail Transit Vehicle,Chongqing Jiaotong University,Chongqing 400074,China)
出处
《组合机床与自动化加工技术》
北大核心
2019年第11期78-82,共5页
Modular Machine Tool & Automatic Manufacturing Technique
基金
重庆市基础与前沿研究计划项目(cstc2016jcyjA0526)
重庆市教委科学技术研究项目(KJ1600519)
重庆市社会事业与民生保障科技创新专项项目(cstc2017shmsA30016)
关键词
小波分解
样本熵
遗传优化算法
wavelet decomposition
sample entropy
genetic optimization algorithm
