基于贝叶斯网络的信息融合汽车网络故障诊断方法  

Fault Diagnosis Method of Automotive Network with Bayesian Network-based Information Fusion

作　　者：李宏梅[1,2] 佟为明[1] 程树康[1] 

机构地区：[1]哈尔滨工业大学电气工程及自动化学院,哈尔滨150001 [2]航天科技控股集团股份有限公司,哈尔滨150060

出　　处：《汽车工程》2015年第10期1190-1194,共5页Automotive Engineering

基　　金：国家自然科学基金(51377030)资助

摘　　要：本文中基于贝叶斯网络的信息融合,开展汽车网络故障诊断方法的研究。首先分析了汽车网络系统的故障类型;其次,针对其故障具有随机性和不确定性的特点,在分析了相关故障诊断方法的基础上,提出了基于贝叶斯网络的信息融合故障诊断方法;最后,结合专家知识和大量统计数据,将所提出的故障诊断方法应用在汽车发动机节点的故障诊断中,证实了该方法有效性。Based on the information fusion of Bayesian network,the fault diagnosis method of vehicle network is studied in this paper.Firstly,the fault types of vehicle network system are analyzed.Then,according to the randomness and uncertainty features of faults,a fault diagnosis method with information fusion based on Bayesian network is proposed after analyzing relevant fault diagnosis methods.Finally,combined with expert knowledge and massive statistical data,the fault diagnosis method proposed is applied to the fault diagnosis of automotive engine nodes,validating the effectiveness of the method proposed.

关 键 词：汽车网络 故障诊断 贝叶斯网络 信息融合 

分 类 号：U463.6[机械工程—车辆工程] U472.9[交通运输工程—载运工具运用工程]