基于VMD-SVD与SSA-SVM的电梯导靴故障诊断  

Fault Diagnosis of elevator guide shoe based on VMD-SVD and SSA-SVM

作　　者：李琨[1] 张久亭 陶然[1,2] 许有才 LI Kun;ZHANG Jiu-ting;TAO Ran;XU You-cai(Faculty of Information Engineering and Automation,Kunming University of Science and Technology,Kunming 650500,China;Special Equipment Safety Inspection Institute in Yunnan Province,Kunming 650228,China)

机构地区：[1]昆明理工大学信息工程与自动化学院,昆明650500 [2]云南省特种设备检测研究院,昆明650228

出　　处：《信息技术》2021年第10期98-104,共7页Information Technology

摘　　要：为有效提取电梯导靴故障信号特征以及提高故障识别率,提出变分模态分解(VMD)、奇异值分解(SVD)与麻雀搜索算法优化支持向量机(SSA-SVM)相结合的电梯导靴故障诊断方法。首先用VMD算法分解电梯导靴振动信号,利用所得分量构建特征矩阵,再应用SVD理论提取特征矩阵的奇异值序列;用麻雀搜索算法(SSA)优化支持向量机相关参数,将奇异序列作为特征值输入到SSA-SVM模型,对导靴故障进行识别。结果表明该方法不仅消除了信号分解时的模态混叠现象且有效提取了故障导靴的特征,相比于EMD结合SVM的方法,导靴故障诊断识别率提高了10%。A fault diagnosis method combining variational mode decomposition(VMD),singular value decomposition(SVD)and sparrow search algorithm optimizer support vector machine(SSA-SVM)is proposed to effectively extract the fault signal features of elevator guide shoe and get a higher diagnostic accuracy rate.Firstly,the vibration signal is decomposed by VMD algorithm,and the obtained model component is used to construct the characteristic matrix.Then the SVD theory is applied to extract singular value sequence of characteristic matrix;Finally,the sparrow search algorithm optimizes the support vector machine parameters,and the sequence of singular values is taken as characteristic parameters input into SSA-SVM classification model for the diagnosis of elevator guide boots fault.The simulation results show that the method not only eliminate model mixing caused by signal decomposition,but also effectively extract the features.Compared with the method of EMD combined with SVM,the fault diagnosis and recognition rate of guide boots improves by 10%.

关 键 词：电梯导靴故障诊断 奇异值分解 变分模态分解 麻雀搜索算法 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置]