摘要
传统经验模态分解(EMD)存在模态混叠,难以充分提取故障特征,原始支持向量机(SVM)、相关向量机(RVM)诊断方法核函数存在选取不灵活、结构复杂导致识别效率低的问题,提出了一种结合变分模态分解(VMD)样本熵和混合布谷鸟改进M-RVM的机械传动电机轴承故障诊断新方法。首先,对故障信号进行VMD分解得到多个子序列;然后,筛选其中的有效分量提取样本熵组成故障特征向量;最后,将特征向量输入基于混合布谷鸟算法优化的M-RVM故障诊断模型,达到对电机运行状态准确识别的目的。仿真结果表明,所提方法实现了电机轴承故障状态的准确诊断。与传统分析诊断方法相比,该方法轴承故障识别诊断性能得到提高,对实际工程应用具有重大意义。
Traditional empirical mode decomposition(EMD) decomposition features modal aliasing,and it is difficult to fully extract fault characteristics,while the original support vector machine(SVM),relevance vector machine(RVM) methods have inflexible kernel function selection and complicated structures,resulting in low identify efficiency,thus a new method for fault diagnosis of mechanical transmission motor bearings based on variational mode decomposition(VMD) sample entropy and hybrid cuckoo improved M-RVM is proposed.Firstly,the fault signal is decomposed by VMD to obtain multiple sub-sequences,then the active components are extracted by the sample entropy to compose the fault feature vector;finally,the feature vector is input to M-RVM fault diagnosis model based on hybrid cuckoo algorithm optimization,to achieve the purpose of accurately recognizing the running state of the motor.The simulation results show that the proposed method contributes to the accurate diagnosis of motor fault.Compared with the traditional analysis and diagnosis methods,the bearing fault diagnosis performance is improved,which is of great significance for practical engineering applications.
作者
路照坭
朱希安
LU Zhaoni;ZHU Xian(College of Information and Communication Engineering,Beijing Information Science and Technology University,Beijing 100101,China)
出处
《自动化仪表》
CAS
2019年第9期46-51,共6页
Process Automation Instrumentation
基金
北京市科技创新服务能力建设提升计划基金资助项目(PXM2018_014224_000012)
关键词
轴承
故障信号诊断
变分模态分解
特征提取
样本熵
改进混合布谷鸟算法
多分类相关向量机
故障分类识别
Bearing
Fault signal diagnosis
Variational mode decomposition
Feature extraction
Sample entropy
Improved hybrid cuckoo search algorithm
Multi-classification relevance vector machine
Fault classification identification
