GA-CFS结合案例推理的轴承故障诊断

Bearing Fault Diagnosis Based on GA-CFS and Case-Based Reasoning

针对轴承故障诊断中知识难以获取的问题,提出了一种GA-CFS(Genetic Algorithm and Correlation-Based Feature Selection,GA-CFS)结合案例推理的轴承故障诊断方法。利用案例推理技术(Case-Based Reasoning,CBR)建立轴承故障案例库进行故障诊断。又针对案例推理技术中案例检索时遇到的属性冗余问题,以及难以人工确定关键属性及其权重的问题,采取GA-CFS方法对属性集合进行筛选,初步确定特征子集,再根据遗传算法确定各个子集中的特征的权重,最后根据特征子集及其权重选取符合要求的最佳特征子集,再用该特征子集构建轴承故障案例库,并通过实验验证了该方法的可行性。

Aiming at the problem that it is difficult to obtain knowledge in bearing fault diagnosis,a method of bearing fault diagnosis is proposed,which is the GA-CFS(Genetic Algorithm and Correlation-Based Feature Selection,GA-CFS) combined Case-Based Reasoning. Case-based reasoning(CBR)is used to establish a case library of faulty bearings for fault diagnosis. Aiming at the problem of case feature attribute redundancy and difficulty in manually determining the key attributes and their weights in CBR,GA-CFS is used to obtain the attribute feature subset,which is initially determined. The GA determines the weight of the features in each subset,and the best feature subset is finally selected,which meets the requirements according to the feature subset and its weight. Then this feature subset is used to build a case library of faulty bearings. The feasibility of the method is validated through experiments.