柴油机故障诊断的局域波神经网络方法研究

针对柴油机表面振动信号的非平稳性，采用局域波法对其进行有效分解获得多个局域波分量，这些分量有效降低了信号的非平稳性，并且包含着原始信号瞬时频率和模糊频带的双重特征信息。然后以局域波分量的特征参数为输入对RBF神经网络进行训练学习，形成网络。这种方法增强了内燃机故障诊断的可靠性和精确，并在实际柴油机故障诊断中得到了有效地应用。

Fault Diagnosis Method for Diesel Engines Using Local Wave Neural Network

In view of the non-stationary and non-linearity characteristics of the vibration signals from the surface of the diesel engine,a new method,which was called local-wave(LW) method,was presented to decompose those signals into a number of intrinsic weighs.These weighs reduced the non-stationary of the signal effectively,and contained the instantaneous frequency and blur frequency,the double character information of the original signals.Then the RBF network was constructed and trained using these weights as inputs.This LW method enhanced the reliability and precision of fault diagnosis,and was used in the fault diagnosis of diesel engines effectively.