An Improved Bayesian Network Inference Algorithm

Abstract

In the on-line fault diagnosis of auto engineer before factory in FAW, we adopt Bayesian network inference to get diagnosis result. To reduce inference complexity, an improved Bayesian network inference algorithm is presented based on graph search strategy under Martelli standard. Through proof, the complexity of the improved algorithm can reduce from exponential level to polynomial level. In experiment, the algorithm has been realized and been compared with expert system method, the experiment shows that the improved algorithm can improve the diagnosis efficiency. The algorithm has been applied in the on-line fault diagnosis of auto engineer before factory in FAW successfully.