基于能量熵特征的挖掘机液压系统故障诊断  

Excavator Hydraulic System Fault Diagnosis Based on Energy Entropy

作　　者：王衡 傅波[1] 田鹏[1] 柳晓东[1] WANG Heng;FU Bo;TIAN Peng;LIU Xiao-dong(School of Mechanical Engineering,Sichuan University,Chengdu 610065,China)

机构地区：[1]四川大学机械工程学院,四川成都610065

出　　处：《液压气动与密封》2020年第10期84-89,共6页Hydraulics Pneumatics & Seals

基　　金：四川省科技计划项目(2018JY0620);川大-泸州战略合作专项资金项目(2014CDLZ-G15)。

摘　　要：液压系统作为挖掘机的核心组成,其工作的可靠性将直接影响主机的性能。在现有较为先进的故障模式判别方法中,选择应用小波包分析计算故障信号的能量熵,构造特征向量输入到支持向量机中进行训练,判定故障的类型。该文融合小波包数据处理技术和支持向量机故障模式识别技术来实现故障的有效判别,完成了液压系统的状态监测及故障诊断系统的搭建。通过测试发现,故障识别的准确率良好,试验结果也验证了支持向量机在处理小样本方面的优势所在。As the core component of excavator,the reliability of hydraulic system will directly affect the performance of the main engine.Among the advanced methods of fault pattern recognition,wavelet packet analysis is used to calculate the energy entropy of fault signal,and eigenvectors are constructed and input to support vector machine for training to determine the type of fault.This essay integrates wavelet packet data processing technology and support vector machine fault pattern recognition technology to realize effective fault identification,and completes the construction of condition monitoring and fault diagnosis system of hydraulic system.The test results show that the accuracy of fault identification is good,and the experimental results also verify the advantages of support vector machine in dealing with small samples.

关 键 词：挖掘机 液压系统 支持向量机 小波包分解 

分 类 号：TH137[机械工程—机械制造及自动化] TH133.33