摘要： 针对轴承故障振动信号特点，提出一种基于离散余弦变换(DCT)、遗传算法(GA)和支持向量机(SVM)的轴承故障诊断方法。利用DCT的能量聚集性在广义频域建立原始特征向量集，运用GA以SVM的最低分类错误率为目标函数建立故障特征向量集，使用SVM完成轴承故障诊断。分别对轴承内圈故障、外圈故障、滚动体故障进行故障诊断，结果表明，该方法能够准确诊断轴承故障。

关键词: 故障诊断, 离散余弦变换, 遗传算法, 支持向量机, 轴承, 振动信号

Abstract: Aiming at the characteristics of bearing fault vibration signal, a scheme for bearing fault diagnosis based on Discrete Cosine Transform(DCT), Genetic Algorithm(GA) and Support Vector Machine(SVM) is proposed. The energy accumulation of DCT is utilized to build an original feature vector set in generalized frequency domain. GA implements the process of constructing a fault feature vector set, while the least error fault diagnosis rate is the target function of GA. SVM performs bearing fault diagnosis. Bearing inner race fault, outer race fault and rolling element fault are diagnosed respectively. Results show that this scheme can diagnose bearing fault correctly.

Key words: fault diagnosis, Discrete Cosine Transform(DCT), Genetic Algorithm(GA), Support Vector Machine(SVM), bearing, vibration signal

中图分类号: 

• TH113.1