A state searching method for model-based diagnosis

Model-based diagnosis has a great advantage compared to other diagnostic methods. However, there are some problems existed in model-based method like large complexity, inefficient. This paper studies a state searching method using system conflicts to exclude large invalid state space and speed up the search procedure. All the tested candidate states are generated in best first order and expanded to find best solutions of the system. The simulation results prove the proposed searching method is more effective than the general method.