基于神经网络的柴油发动机故障预测研究  

Research on Fault Predication for Diesel Motor Based on Neural Network

作　　者：康健[1] 左宪章[1] 吴彩华[2] 于晓伟[2] 

机构地区：[1]军械工程学院电气工程系,河北石家庄050003 [2]军械工程学院计算机工程系,河北石家庄050003

出　　处：《计算机测量与控制》2006年第8期987-989,共3页Computer Measurement &Control

基　　金：总装备部司令部基金资助项目(2005HZ4309002)

摘　　要：为了保证设备安全可靠地运行,需要对设备进行故障预测,及时消除故障;探讨了故障预测技术的应用及神经网络预测原理和建模的方法,给出了神经网络模型在预测过程中的学习算法和改进的神经网络的训练算法,并用于柴油发动机运行中的故障预测,经过多种柴油机的试验证明试验结果满意,对故障的实时检测有重要意义,具有良好的应用前景。In order to make the device run safely and reliably, fault diagnosis of the device is needed. The fault predication technique and the predication theory and the method of neural network are studied. An algorithm of modeling and training is given for the neural network in the course of predication. It has made very pleased result through experiment of many kinds of diesel motor and it is important to real time detecting of fault, and it has good prospects.

关 键 词：神经网络 柴油发动机 故障预测 故障诊断 

分 类 号：TP29[自动化与计算机技术—检测技术与自动化装置]