High-Speed Stochastic Qualitative Reasoning by Model Division and States Composition

Stochastic qualitative reasoning is an effective way to grasp approximate behavior of complex systems such as air conditioning systems . The appropriateness of the stochastic qualitative model can be identified by comparing the behavior derived by reasoning that is represented as the transition of states with the actual measured behavior . Since the states are derived based on all conceivable combinations of rule application, the number of derived states exponentially increases with the size of the qualitative model . If the model is large, reasoning cannot be completed in real time . This paper proposes a method of high-speed stochastic qualitative reasoning . In this method, model division and states composition are introduced . First, the partial models are constructed by dividing the entire model . Next, reasoning is executed in each partial model . Finally, the states in the entire model are generated by composing derived states in each model . By this method, the number of derived states is reduced, and the reasoning time is shorter than for the one by the previous method and reasoning can finish regardless of model size . This method was applied to an actual air conditioning system . It was confirmed that stochastic qualitative reasoning with model division and states composition derived the same results as the previous method did .