基于自回归预测的故障报警方法研究  

Research on fault alarming method based on auto-regression forecasting

作　　者：柴令华[1] 陈小虎[1] 孟凯[1] 何庆飞[1] 

机构地区：[1]第二炮兵工程学院501教研室,陕西西安710025

出　　处：《机械》2010年第11期6-8,共3页Machinery

摘　　要：工业生产过程由于设备的复杂性,系统的安全性和可靠性日益受到人们的重视。研究了工业生产中的故障报警方法技术,通过建立自回归模型,利用拉依达法确定系统的报警阈值。以齿轮泵泵壳振动信号作为研究对象,对其常见的四种故障进行了实验验证,并在液压试验台上进行实验。结果表明,该方法故障报警率较高,能够有效地对齿轮泵故障进行报警。由于报警阈值利用每一次设备运行时的数据进行动态调整,可以实现对设备故障进行及时有效地报警。Because of the complexity of equipment in modem industrial production course, systematic safety and reliability get more attention by people. In this paper, the fault alarming method in industrial production was studied, through establishing auto-regression model, confirmed the alarming threshold value of the produce system by the way of layida. Using the vibration signal of gear pump as research object, the experiments validate its four common faults. The experiment on the platform of the hydraulic proved that the method had higher ratio to the fault alarming, could give alarming of gear pump effectively. Because of the alarming threshold value can carry through dynamic redress by using the data in each equipment operation, the method could realize fault alarming to equipment in time and effectively.

关 键 词：故障预测 自回归模型 动态报警 

分 类 号：TH325[机械工程—机械制造及自动化]