重庆理工大学学报（自然科学版）2016，Vol.30Issue(3)：34-39,6.DOI:10.3969/j.issn.1674-8425(z).2016.03.006
基于高斯混合模型与改进网格搜索法的轴承故障诊断
Bearing Fault Diagnosis Based on Gauss Mixture Model and the Improved Grid Search Method
摘要
Abstract
The paper carried out the study of the method of bearing fault diagnosis under different conditions and combined the Gaussian mixture model with the improved grid search method to improve the diagnosis accuracy of complex bearing fault. The paper paragraphed the data appropriately and fitted the segment data using the mixed Gauss distribution,extracting statistical characteristics as the fault feature index,and then using common grid search method and the improved grid search method respectively,we had parameter optimization;finally,the paper used the support vector machine as a classifier for bearing fault diagnosis and compared the accuracy of two kinds of optimization algorithm. The results show that the accuracy is higher by using the proposed fault diagnosis method.关键词
高斯混合模型/统计特征量/支持向量机/改进网格搜索法/故障分类Key words
Gauss mixture model/statistical characteristics/support vector machine/improved grid search method/fault classification分类
机械制造引用本文复制引用
陈远帆,李舜酩..基于高斯混合模型与改进网格搜索法的轴承故障诊断[J].重庆理工大学学报（自然科学版）,2016,30(3):34-39,6.基金项目
中央高校基本科研业务费专项资金资助项目（NZ2015103）；机械结构强度与振动国家重点实验室开放课题资助项目 （）
