中文摘要：

针对液压系统性能参数的退化特点,提出了一种小波包变换和隐马尔科夫模型（HMM）相结合的液压系统故障预测方法.该方法对采集的振动信号进行小波包变换,提取能量特征,分别使用正常状态,经过轻度退化状态、中度退化态和高度退化态,最终达到故障状态的全过程数据训练HMM,建立性能评估模型,然后进行模式识别,实现液压系统的故障预测.最后,通过试验研究,验证了所提出的方法的可行性和有效性.

英文摘要：

Because the parameters of the performance of a hydraulic system have degradation,we propse a method for predicting its faults,which combines the wavelet packet transform（WPT） with the hidden markov model（HMM）.The WPT was used to extract features from vibration signals.HMMs were trained respectively using the data under normal condition,minor degradation condition,moderate degradation condition,high degradation condition and fault condition.Moreover,the trained HMMs were used to pattern recognition and hydraulic system failure prediction.Finally the experimental results show that the proposed methodology is feasible and effective.