摘要
针对滚动轴承振动信号故障难以识别的问题,创建一种应用改进鲸鱼优化算法(IWOA)优化支持向量机(SVM)的故障诊断模型。首先将轴承振动信号特征通过VMD的方式提取;其次,为改进鲸鱼优化算法,采取精英反向学习策略增强种群的广泛性,选用非线性因子并加入随机扰动策略增强探索能力;通过4组基准测试函数,将IWOA与4种优化算法对照分析,验证了此改进算法的优越性;最后,将SVM的惩罚参数和核函数参数放入IWOA中,构建IWOA-SVM故障分类模型。故障诊断的结果表明,用IWOA-SVM分类模型在故障诊断中拥有更好的效果,准确率达到100%。
Regarding the issue that the fault of rolling bearing vibration signal is arduous to identify,a fault diagnosis model using improved whale optimization algorithm(IWOA)to optimize support vector machine(SVM)is established.Firstly,the bearing vibration signal features are extracted by means of VMD.Secondly,to improve the whale optimization algorithm,an elite reverse learning strategy is employed to strengthen the scalability of the population,a nonlinear factor is selected and a random perturbation strategy is added to enhance the exploration capability.The superiority of this improved algorithm is verified by four sets of benchmark test functions,and IWOA is analyzed against four optimization algorithms.Finally,the penalty parameters of SVM and kernel function parameters are put into IWOA to construct IWOA-SVM fault classification model.According to the results of fault diagnosis,the IWOA-SVM classification model has better results in fault diagnosis,and the accuracy rate reaches 100%.
作者
纪佳呈
张金萍
JI Jiacheng;ZHANG Jinping(School of Mechanical and Power Engineering,Shenyang University of Chemical Technology,Shenyang 110142,China)
出处
《机械工程师》
2023年第8期50-53,共4页
Mechanical Engineer
