基于FARX模型的工程机械液压系统故障特征提取方法研究  

Fault Feature Extraction of Construction Machinery Hydraulic System Based on FARX Model

作　　者：贺湘宇[1] 何清华[2] 

机构地区：[1]长沙理工大学汽车与机械工程学院,长沙410004 [2]中南大学机电工程学院,长沙410083

出　　处：《机械科学与技术》2011年第12期2008-2012,共5页Mechanical Science and Technology for Aerospace Engineering

基　　金："863"国家高技术研究发展计划项目(2003AA430200);湖南省教育厅项目(09C075);长沙理工大学"湖湘学者"项目(200807)资助

摘　　要：为了提高工程机械的可靠性水平和智能化程度,研究了FARX(Fuzzy Auto-Regressive witheXtra inputs)模型非线性特征及其故障特征参数的提取过程,提出了基于FARX模型的工程机械液压系统故障特征提取方法。相对于普通的ARX模型,FARX模型具有的非线性特征提取能力。实验分析中,该方法以目标故障特征为分类参考,使用FCM(Fuzzy Clustering Means)分类器对故障特征进行分类,判断系统的故障状态。5类典型液压系统故障作为目标故障对故障诊断方法进行验证。基于FARX模型的故障诊断方法准确地诊断出了90%以上的系统故障。In order to improve reliability and intelligence level of construction machinery, the FARX （ Fuzzy AutoRegressive with eXtra inputs） model structure is established and a fault feature extraction approach is proposed for the hydraulic system of construction machinery based on FARX model. Compared with classical ARX model, FARX model is capable to extract nonlinear features of hydraulic system. On the basis of target fault features, FCM （ Fuzzy Clustering Means） is served as fault classifier and the output of FCM is the result of fault diagnosis. Five typical faults of hydraulic system were used to test the fault diagnosis approach proposed in this paper. Experimental results show that 90% of the test faults were correctly identified, and the fault diagnosis approach proved feasible and effective on construction machinery hydraulic system.

关 键 词：液压系统 工程机械 故障诊断 模糊逻辑 有源自回归(ARX)模型 

分 类 号：TP206.3[自动化与计算机技术—检测技术与自动化装置]