Reasoning in fuzzy production systems when input information is incomplete

This paper presents a new incomplete reasoning technique in fuzzy production systems when the given input information is incomplete. Incomplete inputs normally infer no conclusive results from the rule bases. However, possible conclusions may be obtained if minimum number of missing facts are assumed to be true. Under fuzzy Petri net formalism for fuzzy production systems, the reachability set of the given incomplete input contains the potential conclusions. The potential missing inputs can be inferred by using fuzzy backward reasoning scheme on the reachability set. The final possible conclusions are drawn as those in the reachability set that require minimum number of missing facts assumed the be true. Since real decisions are typically made under incomplete input information, our algorithm provides another realistic application of fuzzy production systems.