摘要
电液伺服阀是电液伺服控制中的关键元件,其性能关系到整个伺服系统的控制精度和响应速度。当前,伺服阀的故障诊断仍以离线为主,缺乏在线诊断的有效手段。根据伺服阀的工作特性,提出一种反映伺服阀动态特征的状态信号选取方法;通过对伺服阀阀芯开口度进行时频联合分析,结合所选取的伺服阀特征参数,提取出反映伺服阀动态特征信息的特征向量;采用粗糙集理论对特征量进行约简以提高在线诊断效率。基于人工神经网络的伺服阀性能在线诊断的实验结果表明:所提取的特征向量能够准确反映伺服阀动态特征信息,有效判断伺服阀的异常状态,为电液伺服阀的在线故障诊断提供了参考。
Servo valve is one of the most important key components in hydraulic servo control systems.Its performance determines the system's control precision and response speed.Currently,fault diagnosis of servo valve is mainly based on off-line.A new method of selection of status signals was presented to describe the dynamic features of the valve based on the operating characteristics.The time-frequency analyses of opening degree of the valve core were made.The analysis results were combined with the other valve's characteristic parameters to form the feature vector to reveal the dynamic characteristic of the servo valve.The feature components were reduced based on the rough sets theory to improve the online diagnosis efficiency.The experimental results show that using the proposed method,the dynamic characteristics information of the valve can be effectively extracted and the abnormal states of the valves can be revealed.It provides reference for the online fault diagnose of the servo valve.
出处
《机床与液压》
北大核心
2011年第11期15-19,23,共6页
Machine Tool & Hydraulics
基金
机械系统与振动国家重点实验室开放基金资助项目(MSV-2009-16)
关键词
伺服阀
特征提取
在线诊断
时频分析
粗糙集
Servo valve
Feature extraction
On-line diagnosis
Time-frequency analysis
Rough sets