A Fuzzy Reasoning Algorithm in Hybrid Causality Diagram

摘要

This paper discussed the reasoning algorithm of hybrid causality diagram with fuzzy method, it map continuous-value into discrete states mainly, and this algorithm can also deal with fuzzy multi-value causality diagram, continuous-value causality diagram. The construction steps of membership functions were developed, the probability of event state was proposed. The calculation method of posterior probability of the interesting variable was given when evidence received, and the corresponding calculation formula is derived. This theoretical construct in this paper provides a new inference engine for inference under two types of uncertainty: fuzziness and randomness. At last, two examples are given which show that the algorithm is effective.