基于神经网络的电控发动机怠速不稳定的故障识别方法

Fault Identification of Electric Control Engine for Unstable Idle Speed Based on Neural Network

以电控发动机怠速不稳定这种十分常见的故障为研究对象 ,系统地介绍了应用神经网络进行故障原因识别的方法。详细地分析了不同网络结构的训练效率和以不同样本形式训练的网络对相同故障症兆的故障原因的识别能力。

In this paper, a fault identification method for unstable idle speed of electric control engine is studied. It introduces systematically how to identify the fault reasons of the unstable idle speed by means of neural network, and analyzes detailedly the training efficiency of different neural network construction and the recognizing ability of different type data samples for same fault reason.