摘要
文章介绍了故障自诊断的步骤,针对自动变速器电控系统的特点,建立了故障征兆与故障原因映射列表,并应用基于模糊神经网络的故障自诊断方法,实现了对自动变速器电控系统故障的模糊诊断;并介绍了模糊神经网络的结构和学习方法,以及如何对训练样本进行预处理;利用Matlab神经网络工具箱作为平台,进行测试样本仿真和试验,结果表明该诊断方法是行之有效的。
This paper firstly introduces the approaches to the design of fault self-diagnosis. Then a list of fault symptom and fault cause is made in allusion to the characteristics of electronic control system of automatic transmission, and the approach of fault self-diagnosis based on fuzzy neural network is applied to realize the fuzzy diagnosis of faults in the mentioned system. The structure of the fuzzy neural network and corresponding learning methods are introduced as well as the ways to pretreat the training samples. Finally, this diagnosis approach is simulated and tested by Matlab＇s Neural Networks Toolbox to demonstrate its validity and practicability.
出处
《合肥工业大学学报（自然科学版）》
CAS
CSCD
北大核心
2010年第4期501-505,518,共6页
Journal of Hefei University of Technology：Natural Science
基金
国家科技部支撑计划重点资助项目(2009BAG12B03)
安徽省自然科学基金资助项目(08040102001)
合肥工业大学科学研究发展基金重点资助项目(105-036413)
关键词
模糊逻辑
神经网络技术
自动变速器电控系统
故障自诊断
fuzzy logic
neural network technique
electronic control system of automatic transmission
fault self-diagnosis
