基于IAGA-SVM的捣固车液压系统故障诊断研究

摘要

针对传统液压系统故障诊断方法受人为因素影响较为严重,故障成因相对复杂等问题.提出一种改进的自适应捣固车液压系统故障诊断方法.首先,从捣固车的车载数据中采集系统抽取出来的故障特征值.其次,将特征值输入支持向量机(SVM)模型中进行训练,同时对核函数和惩罚系数做出优化.最后,应用自适应支持向量机建立从特征向量到故障模式之间的映射,最终做到对液压系统的故障诊断.结果可得,此方法可以准确高效地诊断出故障类型,证明了此方法的实用价值.此外,经过与GA-SVM以及AGA-SVM的对比剖析,表明了IAGA-SVM方法在故障诊断领域中的卓越性.%In view of the traditional hydraulic system fault diagnosis methods are affected by human factors and the causes of the faults are relatively complex,a fault diagnosis method of tamping machine hydraulic system of an improved adaptive is proposed.First,the fault diagnosis eigenvalue extracted from the vehicle data acquisition system of tamping machine were collected.Second,the eigenvalue input support vector machine (SVM) model was trained.Meanwhile,kernel functions and the penalty coefficient were optimized.Moreover,the adaptive support vector machine was applied to establish the mapping between the feature vector and the fault model,and finally the fault diagnosis of the hydraulic system was done.The results show that the method can quickly and accurately diagnose the fault of rolling bearing,and verify the validity and stability of this method.In addition,through the comparative analysis with GA-SVM and AGA-SVM,it shows the superiority of the IAGA-SVM method in the intelligent fault diagnosis application.