| 摘要 |
PURPOSE:To provide the fuzzy set processing system which can simplify the constitution of a neural network in a neuro-fuzzy fusion system. CONSTITUTION:By a fuzzy set processing means 1, a fuzzy model is generated by extracting an input/output relation of an object 3 by a format of a fuzzy rule 12 having an antecedent part proposition and a consequent part proposition, and a membership function 13. In accordance with the fuzzy model, a weight value of coupling between neurons 20 and a threshold are set by a neural network 2. In the fuzzy set processing means 1, a proposition logical expression converting part 15 for converting a logical expression of the antecedent part proposition of the fuzzy rule 12 to a logical expression of a products-sum standard type is provided, and in accordance with the logical expression of the products-sum standard type converted by the proposition logical expression converting part 15, the neural network 2 is constituted. |
