最小二乘支持向量机在电梯故障诊断中的应用

摘要

Aimed at the puzzle of being difficult to extract the fault features and less in fault data sample,the paper proposed a sort of diagnosis method of elevator fault based on least square support vector machine(LS-SVM).By means of wavelet packet analysis and LS-SVM,it firstly extracted the elevator fault data samples by wavelet packet analysis,and then identified the elavator fault based on their characteristics by use of LS-SVM.The test data of actual experiment demonstrsted that it has better performance in diagnosis for jerk fault by the proposed method.The research result shows that it is excellent in diagnosis performance.%针对电梯故障诊断中特征提取困难和故障样本数量少问题,提出了应用小波包分解和最小支持向量机(LS-SVM)相结合进行电梯急停智能故障诊断的方法.借助小波包分解,该方法首先提取电梯轿厢振动信号作为特征向量,然后利用LS-SVM分类模型对故障进行辩识.实验证明,小波包与LS-SVM相融合的故障诊断与识别技术可发挥两者的优势,该方法对电梯急停故障的诊断具有较好的诊断效果.