基于RBF神经网络的发动机怠速抖动故障诊断

针对难以获取发动机怠速抖动的精确数学模型,文章提出一种基于RBF神经网络的模型逼近算法。该算法利用RBF神经网络良好的模型逼近能力,对造成发动机怠速抖动的几种典型故障原因进行了诊断。实验结果表明该算法具有良好的逼近效率和诊断效果。

Engine Idle Wobbling Fault Diagnosis Based on RBF Neural Network

For the accurate mathematical model of engine idle wobbling was difficult to be established, a new approximation algorithm based on RBF neural network was proposed. In this algorithm,the RBF neu-ral network was used to approximate the model approximation ability, and some typical faults of engine idle wobbling were diagnosed. Experimental results show that the proposed algorithm has good efficiency and diagnostic efficiency.