工程机械液压系统状态监测的神经网络模式识别方法  

Nerval Network Pattern Recognition Method for Condition Monitoring of Construction Machinery Hydraulic Systems

作　　者：石红雁[1] 许纯新[1] 瞿爱勤[1] 

机构地区：[1]吉林大学

出　　处：《工程机械》2004年第2期39-41,共3页Construction Machinery and Equipment

摘　　要：将模式识别技术引入到工程机械液压系统的状态监测,以提高监测的可靠性。首先介绍了三种主要的模式识别方法,在此基础上以阀控液压缸系统为例阐述了如何利用神经网络模式识别技术对液压泵异常状态进行监测。结果表明,本文提出的方法可以快速、准确地监测出液压泵的异常状态。In order to increase the reliability of monitoring, pattern recognition technology was adopted in the condition monitoring of construction machinery hydraulic systems. Three main pattern recognition methods were introduced firstly. On the basis of that, how to monitor the abnormal conditions of hydraulic pumps with nerval network pattern recognition technology was explained by an example of valve-controlled hydraulic pump system. The result shows that the abnormal conditions of hydraulic pumps can be monitored fast and accurately with the method discussed in the article.

关 键 词：工程机械 液压系统 状态监测 神经网络 模式识别 可靠性 

分 类 号：TH137[机械工程—机械制造及自动化]