基于粗糙集理论的柴油机燃油系统故障诊断研究

Fault diagnosis study of diesel engine injection system based on rough sets theory

针对柴油机燃油系统的故障种类多的特点利用小波方法对燃油波形进行分析,提取时域及频域的故障特征参数,并使用粗糙集理论对这些故障特征参数进行约简,达到简化故障识别的神经网络的结构、加快辨识速度的目的.最后用RBF神经网络对所模拟的各类故障进行辨识,证实了粗糙集理论在柴油机燃油系统故障诊断中应用的可行性.

In view of many kinds of faults of diesel engine fuel injection system, this paper analyzed the fuel pressure wave shape using wavelet theory, picked up the typical fault parameters in both time domain and frequency domain and used rough sets theory to reduce them so as to simplify the neural network structure of fault identification and speed the rate of identification, finally used RBF neural networks to identify all kinds of faults simulated here. The results have shown that it would be feasible to apply the theory in the fault diagnoses of diesel engine fuel injection system.