Research on Fault Diagnosis for Vehicles Based on Bayesian Network Model Constructing
Hou Yong-yan · Jisuanji fangzhen · 2011
To resolve the uncertainty and modeling issues in fault diagnosis for vehicles,a fault diagnosis fusion system architecture based on Bayesian network model constructing is proposed.A fault diagnosis algorithm based on Bayesian network constructing is also advanced.This fault diagnosis approach handles with the uncertain representation and reasoning by exploiting the learning and probabilistic inference abilities of Bayesian network.Moreover,the provided system can realize the self adaptation via the fusion of domain prior knowledge and the distribution of real time data from the sensor system,which is successfully applied in vehicle fault diagnosis.Experimental results demonstrate that the proposed approach can supply the accurate and reliable decision-making support in the fault diagnosis.