基于LF-GWO优化FKCA模型的齿轮箱故障诊断研究  

Gearbox Fault Diagnosis Based on LF-GWO Optimized FKCA Model

作　　者：袁荷伟[1] 李高磊[1] 袁黎[2] 张强 YUAN He-wei;LI Gao-lei;YUAN Li;ZHANG Qiang(He'nan Vocational and Technical College of Communications,College of Automobile,He'nan Zhengzhou 450000,China;Hohai University,College of Civil Engineering and Communications,He'nan Zhengzhou 450000,China;He*nan Intelligent Hydraulic Equipment Co.,Ltd.,He,nan Zhengzhou 450000,China)

机构地区：[1]河南交通职业技术学院汽车学院,河南郑州450000 [2]河海大学土木与交通学院,河南郑州450000 [3]河南睿智液压设备有限公司,河南郑州450000

出　　处：《机械设计与制造》2023年第4期239-242,共4页Machinery Design & Manufacture

基　　金：河南省高等学校青年骨干教师培养计划项目(2017GGJS242)。

摘　　要：采用莱维飞行策略对灰狼优化算法进行了优化,显著提升了算法的初期搜索性能。建立了LF-GWO算法求解FKCA模型,并给出了故障诊断步骤。通过实验验证结果表明:经过800次迭代计算处理时,LF-GWO寻优形成的最小错误率2%,以LF-GWO/FKCA诊断测试集时获得了98%的正确率,只对点蚀与磨损的故障类型存在错误判断各一处情况。相比较FKCA和BP方法,采用LF-GWO/FKCA方法则可以将无标签缺齿数据归为第4类,从而实现与已知故障类型的区分,达到了更高的正确率,实际测试正确率为98%,显著提升故障诊断正确率。对齿轮箱的故障进行诊断仿真显示本文设计的诊断方法可以达到很低的错误率,表现出了优异的故障诊断性能。Levy flight strategy was used to optimize the grey Wolf optimization algorithm,which significantly improved the initial search performance of the algorithm.LF-GW 0 algorithm is established to solve the FKCA model,and the steps of fault diagnosis are given.The experimental verification results show that after 800 iterations,the minimum error rate formed by LF-GWO optimization is 2%,and the accuracy rate is 98%when LF-GWO/FKCA is used to diagnose the test set.There is only one wrong judgment for the fault types of pitting and wear.Compared with the FKCA and BP methods,the LF-GWO/FKCA method can classify the number of label-free tooth missing data into the fourth category,thus realizing the distinction from the known fault types,and achieving a higher accuracy rate than the GWO-FKCA method.The actual test accuracy rate is,8%,significantly improving the accuracy rate of fault diagnosis.The simulation of gearbox fault diagnosis shows that the method designed in this paper can achieve a very low error rate and shows excellent fault diagnosis peifarmance.

关 键 词：灰狼优化算法 模糊核聚类 齿轮箱 故障诊断 错误率 

分 类 号：TH16[机械工程—机械制造及自动化] TH137