A Study on Particle Fuzzy Reasoning Method

General fuzzy reasoning methods, e.g. Mamdani's method and the same kinds, can deal with fuzzy input and fuzzy output. But, its fuzzy outcomes are liable to be complex shape membership function. Therefore, its representative value is generally obtained by defuzzification technique, and is used as a reasoning result. This paper proposes the Particle Fuzzy Reasoning Method that deals with fuzzy input and fuzzy output, too. This method makes the input fuzzy set discrete values, and considers them to be particles. Each particle is used for general fuzzy reasoning as an input. The reasoning results from the particles are united into one. If a fuzzy number is given as an input, the fuzzy outcome becomes a fuzzy number.