基于QPSO-SVM的轴承故障诊断方法  

Method of Rolling Bearing Fault Diagnosis based on SVM with QPSO

作　　者：杨光春[1] 蹇清平[1,2] 

机构地区：[1]攀枝花学院机械工程学院,四川攀枝花617000 [2]西南石油大学机电工程学院,四川成都610500

出　　处：《机械传动》2014年第8期134-138,共5页Journal of Mechanical Transmission

摘　　要：轴承作为旋转机械中应用广泛的一类支撑部件,其故障将严重影响设备的安全运行,为了实现对轴承故障的有效诊断,提出一种量子粒子群优化(Quantum particle swarm optimization,PSO)支持向量机(Support vector machine,SVM)的故障诊断模型,首先采用经验模式分解(Empirical mode decomposition,EMD)方法将故障信号分解为多个固有模态分量(Intrinsic mode function,IMF)之和,其次,提取表征轴承故障特征的IMF分量能量构造特征向量,最后采用QPSO优化的SVM模型对故障模式进行识别。实验结果表明,所提出的轴承故障诊断方法具有自适应提取轴承故障特征和高精度的自适应诊断能力。Due to the importance of rolling bearing as one of the most widely used in rotating machines,bearing failures have adverse effects on the safe operation of the equipment,in order to diagnosing the fault of rolling bearing effectively,a fault diagnosis model of support vector machine（SVM）optimized by quantum particle swarm optimization（QPSO）algorithm is proposed.First,fault vibration signals are decomposed into several intrinsic mode functions（IMFs）using empirical mode decomposition（EMD）method,then the instantaneous amplitudes of the IMFs that have the fault characteristics are extracted and regarded as the features vector,finally the SVM model optimized by QPSO is used for the failure mode identification.The experimental results indicate that the proposed bearing fault diagnosis method has good capability for adaptive features extraction as well as high fault diagnostic accuracy.

关 键 词：量子粒子群 支持向量机 参数优化 故障诊断 EMD分解 

分 类 号：TH165.3[机械工程—机械制造及自动化] TH133.3