IPSO-RVM的轴承故障识别方法

Research on Bearing Fault Diagnosis Method Based on IPSO-RVM

针对轴承故障诊断过程中传统粒子群优化支持向量机所存在的分类效果较差以及传统粒子群寻优不准确的问题,提出改进粒子群算法优化相关向量机的方法。利用自适应的惯性权重和加速因子,使粒子前期的搜索速度更快,后期的收敛速度更快;构建改进粒子群优化相关向量机(IPSO-RVM)模型和改进粒子群优化支持向量机(IPSO-SVM)模型,与粒子群优化支持向量机(PSO-SVM)模型进行相互对比实验。经仿真实验验证,IPSO-RVM的分类准确率比IPSO-SVM高5.8%,比PSO-SVM高8.7%;IPSO-RVM的运行时间为比IPSO-SVM与PSO-SVM分别慢0.58 s与4.28 s。综上,与PSO-SVM和IPSO-SVM相比,新方法能够在保证时间运行合理的情况下提高分类准确率。

Aiming at the problems of poor classification effect of support vector machine in traditional particle swarm optimization support vector machine and inaccuracy of traditional particle swarm optimization in bearing fault diagnosis,an improved particle swarm optimization method for correlation vector machine is proposed in this paper.By using the adaptive inertia weight and acceleration factor,the search speed is faster in the early stage and the convergence speed is faster in the late stage.The classification models of improved particle swarm optimization correlation vector machine(IPSO-RVM),improved particle swarm optimization support vector machine(IPSO-SVM)and particle swarm optimization support vector machine(PSO-SVM)were constructed respectively for comparative experiments.The simulation results show that the classification accuracy of IPSO-RVM is 5.8%higher than IPSO-SVM and 8.7%higher than PSO-SVM.The simulation results show that the classification accuracy of IPSO-RVM is 5.8%higher than ipSO-SVM and 8.7%higher than PSO-SVM.The running time of IPSO-RVM is 0.58 s and 4.28 s slower than that of IPSO-SVM and PSO-SVM,respectively.Compared with PSO-SVM and IPSO-SVM,the accuracy of classification can be improved under the condition that the time running is reasonable.