基于GSA-VMD和自适应CNN的滚动轴承故障诊断  

Rolling Bearing Fault Diagnosis with GSA-VMD and Adaptive CNN

作　　者：王亚辉[1] 刘德平[1] 王宇[1] WANG Ya-hui;LIU De-ping;WANG Yu(School of Mechanical and Power Engineering,Zhengzhou University,Zhengzhou 450001,China)

机构地区：[1]郑州大学机械与动力工程学院,郑州450001

出　　处：《组合机床与自动化加工技术》2022年第7期85-89,共5页Modular Machine Tool & Automatic Manufacturing Technique

基　　金：河南省重大科技专项(171100210300-01)。

摘　　要：针对轴承故障诊断中,变分模态分解(VMD)的参数选择与卷积神经网络架构难以确定的问题,研究一种GSA-VMD和自适应CNN的滚动轴承故障诊断方法。首先,采用引力搜索算法(GSA)优化VMD的参数,接着利用优化的VMD分解轴承的振动信号得到若干模态分量;然后,将模态分量与振动信号结合构建特征矩阵,作为自适应CNN的输入;最后,自适应CNN采用粒子群算法(PSO)解决CNN架构难以确定的问题,适应性地构建CNN故障诊断模型,判断轴承的故障类型。实验结果表明:与ANN、CNN-SVM、WDCNN、GA-CNN诊断方法对比,所提方法 准确率更高、稳定性好、适应性强。Aiming at the problem that the parameter selection of variational modal decomposition(VMD)and the convolutional neural network(CNN)architecture are difficult to determine in bearing fault diagnosis,a fault diagnosis method for rolling bearing fault diagnosis based on GSA-VMD and adaptive CNN was studied.Firstly,the gravitational search algorithm(GSA)is used to optimize the parameters of the VMD,and then the optimized VMD is applied to decompose the vibration signal of the bearing to obtain several modal components,which are combined with the vibration signal to construct a feature matrix as the input of the adaptive CNN.Finally,adaptive CNN utilizes particle swarm optimization(PSO)to solve the problem of the determination of the CNN architecture and construct adaptively a CNN fault diagnosis model to identify the type of bearing fault.The experimental results showed that compared with ANN,CNN-SVM,WDCNN,GA-CNN diagnosis methods,the proposed method has high accuracy,good stability and strong adaptability.

关 键 词：故障诊断 引力搜索算法 变分模态分解 粒子群算法 自适应卷积神经网络 

分 类 号：TH133.3[机械工程—机械制造及自动化] TG66[金属学及工艺—金属切削加工及机床]