Probability based vehicle fault diagnosis: Bayesian network method

被引:59
作者
Huang, Yingping [1 ]
McMurran, Ross [1 ]
Dhadyalla, Gunwant [1 ]
Jones, R. Peter [2 ]
机构
[1] Univ Warwick, Warwick Mfg Grp, IARC, Coventry CV4 7AL, W Midlands, England
[2] Univ Warwick, Sch Engn, Coventry CV4 7AL, W Midlands, England
关键词
fault diagnosis; Bayesian belief network; modelling; automotive;
D O I
10.1007/s10845-008-0083-7
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Fault diagnostics are increasingly important for ensuring vehicle safety and reliability. One of the issues in vehicle fault diagnosis is the difficulty of successful interpretation of failure symptoms to correctly diagnose the real root cause. This paper presents an innovative Bayesian Network based method for guiding off-line vehicle fault diagnosis. By using a vehicle infotainment system as a case study, a number of Bayesian diagnostic models have been established for fault cases with single and multiple symptoms. Particular considerations are given to the design of the Bayesian model structure, determination of prior probabilities of root causes, and diagnostic procedure. In order to unburden the computation, an object oriented model structure has been adopted to prevent the model from overly large. It is shown that the proposed method is capable of guiding vehicle diagnostics in a probabilistic manner. Furthermore, the method features a multiple-symptoms-orientated troubleshooting strategy, and is capable of diagnosing multiple symptoms optimally and simultaneously.
引用
收藏
页码:301 / 311
页数:11
相关论文
共 10 条
[1]  
FORAN T, 2005, 2005011444 SAE
[2]   An expert system for engine fault diagnosis: development and application [J].
Gelgele, HL ;
Wang, KS .
JOURNAL OF INTELLIGENT MANUFACTURING, 1998, 9 (06) :539-545
[3]   DECISION-THEORETIC TROUBLESHOOTING [J].
HECKERMAN, D ;
BREESE, JS ;
ROMMELSE, K .
COMMUNICATIONS OF THE ACM, 1995, 38 (03) :49-57
[4]  
HUANG Y, 2006, P 8 INT S ADV VEH CO, P469
[5]   Construction of a Bayesian network for mammographic diagnosis of breast cancer [J].
Kahn, CE ;
Roberts, LM ;
Shaffer, KA ;
Haddawy, P .
COMPUTERS IN BIOLOGY AND MEDICINE, 1997, 27 (01) :19-29
[6]   A Bayesian belief network-based advisory system for operational availability focused diagnosis of complex nuclear power systems [J].
Kang, CW ;
Golay, MW .
EXPERT SYSTEMS WITH APPLICATIONS, 1999, 17 (01) :21-32
[7]   Using Bayesian belief networks to predict the reliability of military vehicles [J].
Neil, M ;
Fenton, N ;
Forey, S ;
Harris, R .
COMPUTING & CONTROL ENGINEERING JOURNAL, 2001, 12 (01) :11-20
[8]   Building large-scale Bayesian networks [J].
Neil, M ;
Fenton, N ;
Nielsen, L .
KNOWLEDGE ENGINEERING REVIEW, 2000, 15 (03) :257-284
[9]   Computer-assisted diagnosis of breast cancer using a data-driven Bayesian belief network [J].
Wang, XH ;
Zheng, B ;
Good, WF ;
King, JL ;
Chang, YH .
INTERNATIONAL JOURNAL OF MEDICAL INFORMATICS, 1999, 54 (02) :115-126
[10]   Monitoring and diagnosis of a multistage manufacturing process using Bayesian networks [J].
Wolbrecht, Eric ;
D'Ambrosio, Bruce ;
Paasch, Robert ;
Kirby, Doug .
Artificial Intelligence for Engineering Design, Analysis and Manufacturing: AIEDAM, 2000, 14 (01) :53-67