摘要
针对液压系统中的一种典型故障——油缸泄漏的故障诊断,提出通过监测油缸压力信号,提取时域组合特征并利用神经网络作为故障分类器的诊断方法。该诊断方法首先提取了油缸压力信号的时域组合特征作为特征向量,然后输入到神经网络分类器中进行故障的识别和分类。实验结果表明:该诊断方法能有效识别无泄漏、轻微泄漏、严重泄漏的3种状态,是液压系统故障诊断行之有效的方法。
Aiming at leakage of hydraulic cylinder as a typical failure in hydraulic system,a fault diagnosis approach to leakage of hydraulic cylinder based on monitoring pressure signal,extracting time domain feature and using NN（Neural Network）as a fault analyzer was presented.According to the method,the time domain feature was extracted from the pressure signal and constituted the eigenvectors,then these eigenvectors were input into NN analyzer to identify and classify faults.The experimental results show that three modes of no leakage,slighter leakage and heavy leakage are efficiently identified by this approach which can be used efficiently in the fault diagnosis of hydraulic system.
出处
《机床与液压》
北大核心
2012年第9期158-160,163,共4页
Machine Tool & Hydraulics
关键词
油缸泄漏
故障诊断
神经网络
Leakage of hydraulic cylinder
Fault diagnosis
Neural network
