现代电子技术2016，Vol.39Issue(20)：146-148,153,4.DOI:10.16652/j.issn.1004-373x.2016.20.037
基于PNN神经网络的电控发动机故障诊断
PNN-based fault diagnosis of electrically-controlled engine
摘要
Abstract
The probabilistic neural network (PNN) is a feedforward neural network,and has strong ability to classify the nonlinear patterns. The method of using PNN to diagnose the fault of the engine′s electrically⁃controlled system is proposed. The PNN and its working principle are introduced. The electrically⁃controlled system of the Elantra car engine is taken as the re⁃search object. The Kinder KT600 fault diagnosis instrument is used to collect the failure data flow of the engine while the engine is kept in idle speed. The PNN is used to establish the diagnosis model. The network diagnosis model was verified. The diagno⁃sis results are completely correct,and the training speed is very fast,which show that the PNN has good generalization ability, and has a certain practical value.关键词
PNN神经网络/发动机/电控系统/故障诊断Key words
PNN neural network/engine/electronic controlled system/fault diagnosis分类
信息技术与安全科学引用本文复制引用
巴寅亮,王书提,李春兰,郭增波,加克乌云才次克..基于PNN神经网络的电控发动机故障诊断[J].现代电子技术,2016,39(20):146-148,153,4.基金项目
国家自然科学基金 （）
