摘要
针对转子故障信号中存在非线性、非平稳性的特点,提出采用变分模态分解算法(Variational Mode Decomposition,VMD)对不同程度碰摩信号进行分解,并针对VMD中模态分量个数K、惩罚因子α两个重要参数需人为设定的问题,提出采用粒子群优化算法(Particle Swarm Optimization,PSO)对其进行优化,提高其信噪比,从而提取出更为准确的故障特征;采用支持向量机(Support Vector Machine,SVM)方法对转子碰摩程度进行识别,针对SVM模型中惩罚参数C和核函数参数σ的选取问题,采用PSO进行优化,以建立最优模型,提高识别准确率。通过实验验证,该方法能有效提取出不同程度转子碰磨信号的故障特征,且具有较高的识别率和实际应用价值。
In view of the nonlinear and non-smoothness of the rotor fault signal,the method of using the variable mode decomposition algorithm(Variational Mode Decomposition,VMD)to decompose the different degrees of the mojo signal is proposed.And for the two important parameters(the model component saw K in VMD and the penalty factor alpha)need to be artificially set,the use of particle group optimization algorithm(Particle Swarm Optimization,PSO)to optimize it is proposed,to improve its signal-to-noise ratio,so as to extract more accurate fault characteristics.The support vector machine(Support Vector Machine,SVM)method is used to identify the degree of rotor knock-up.For the SVM model to punish parameters C and nuclear function parameters,PSO optimization is used to establish the optimal model and improve the accuracy of recognition.The experiment verifies that,the method can effectively extract the fault characteristics of different degrees of rotor grinding signal,and has a high erasing rate and certain practical application.
作者
魏永合
姜庆涛
曹怀
WEI Yonghe;JIANG Qingtao;CAO Huai(Shenyang Ligong University,Shenyang 110159,China)
出处
《沈阳理工大学学报》
CAS
2020年第4期42-47,81,共7页
Journal of Shenyang Ligong University
关键词
转子碰摩程度
变分模态分解
支持向量机
粒子群优化算法
rotor rubbing degree
variational mode decomposition
support vector machines
particle swarm optimization
作者简介
魏永合(1971—),男,教授,博士,研究方向:先进制造技术、企业流程管理、设备管理和制造业信息化技术。
