基于RBF神经网络的柴油机故障诊断

Fault Diagnosis of Diesel Engines Based on RBF Neural Network

神经网络模式识别的实时性和鲁棒性使得它成为故障诊断的常用方法。本文首先介绍了RBF神经网络的构成和特性,然后将柴油机的振动信号和油管压力信号作为特征参数,运用RBF神经网络对供油系统的3种故障进行诊断分析。实践表明,RBF神经网络用于多征兆机械系统的故障诊断是有效、可行的。

The pattern recognition based on neural network is often used as the method of fault diagnosis because of its real-time and robustness. In this paper, the characteristic and configuration of RBF neural network are introduced. Vibration signals and fuel pressure signals of diesel engines are used as characteristics parameters to diagnose and analyze three faults of fuel supply system by use of RBF algorithm. The practice shows that the method of RBF neural network is applicable to fault diagnosis of multi-signs engines.