Qualitative reasoning is one of the powerful methods for the fault diagnosis of the complicated systems. The stochastic qualitative reasoning method, one of qualitative reasoning methods, has proposed. However, it is hard to apply this method to large scale systems because of the enormous generation of behavior patterns.
Efficient qualitative reasoning is proposed in this paper. In this method, the minimum existence probability of the generating behavior patterns is estimated by the Monte-Carlo method. The minimum existence probability is used to depress the number of generating behavior patterns. In addition, the improbable behavior patterns in the future are pruned off by future real measured values patterns. This method can be applied to the diagnosis for the large scale systems.