Abstract: Aiming at the difficulty in extracting feature from pressure signal in fault diagnosis for leakage of hydraulic chamber, a fault diagnosis approach based on monitoring pressure signal, wavelet energy feature and BP network is proposed. According to the method, the enegry of different frequency bands after wavelet decomposition constitutes the eigenvectors; these eigenvectors are input into BP network to identify faults. The experimental results show that three modes of no leakage, slighter leakage and heavy leakage are correctly identified and it can be used in the leakage fault diagnosis of hydraulic chamber.

Key words: leakage of hydraulic chamber, fault diagnosis, wavelet energy feature, Back Propagation（BP） network

摘要： 针对液压油缸泄漏故障诊断中压力信号特征提取的难题，提出了通过监测压力信号，基于小波变换能量特征和BP网络的故障诊断方法。该诊断方法将压力信号进行小波分解后得到的各频带信号能量作为特征向量，输入到BP网络分类器中进行故障识别和分类。实验结果表明，该诊断方法能有效识别无泄漏、轻微泄漏、严重泄漏的三种状态，是液压油缸泄漏故障诊断行之有效的方法。

关键词: 液压油缸泄漏, 故障诊断, 小波变换能量特征, 反向传播（BP）网络