基于EMD奇异值熵和GASVM的转子系统故障诊断方法

Rotor System Fault Diagnosis Based on EMD Singular Value Entropy and GASVM

提出一种基于经验模态分解(EMD)和奇异值熵的转子系统故障特征提取方法,克服了奇异值分解相空间重构参数难以选择的问题。然后将奇异值和奇异值熵作为故障特征输入到支持向量机(SVM)中,利用遗传算法(GA)对支持向量机进行参数优化,实现了故障的精确诊断。最后通过对转子不平衡、碰摩和不平衡-碰摩耦合3种故障的正确诊断,证明该方法的有效性。

A fault diagnosis approach for rotor system based on empirical mode decomposition( EMD) and singular value entropy methods was proposed,which has the difficulty in selecting phase-space reconstruction parameter in the process of singular value decomposition solved and the singular value and singular value entropy taken as the fault feature to input into the support vector machine( SVM) as well as the genetic algorithm( GA) used to optimize the SVM parameters so as to realize preliminary fault diagnosis. Diagnosing rotor imbalance,rubbing and imbalance-rubbing coupling proves effectiveness of the proposed method.