近似熵在发动机故障诊断中的应用研究

介绍了近似熵的概念、主要特点及其算法，提出了基于近似熵的特征提取方法，并且讨论了近似熵值计算中3个参数的选择原则，并对发动机声信号进行了分析处理。通过对比正常状态与故障状态的近似熵值，根据其变化有效地提取发动机故障特征，从而实现了对发动机状态的监测与诊断。实验结果表明近似熵在分析复杂信号特征方面具有很强的能力，在判别机械设备运行状况方面具有很好的效果，是一种行之有效的新方法。

Application of Approximate Entropy in Fault Diagnosis of Engine

The concept, main characteristic and the algorithm of approximate entropy （ApEn） are introduced firstly. A method of feature extraction based on ApEn is put forward. Then the select rule of three parameters is discussed in calculating the values of ApEn. Furthermore, engine acoustic signal is processed through this method. By comparing with the ApEn value of normal and fault working condition, the fault features of engine are extracted effectively according to the change of ApEn. So condition monitoring and diagnosis of engine is realized. The experiment resuhs also prove that ApEn has good ability in analy- zing the complex signal feature and is an effective method to identify the running condition.