Algorithm for Fuzzy Relation Identification

Fuzzy model identification of a system can be performed either by subjective or objective methods. Subjective methods assume the describing of system operation in statements of type If...Then... and their summarization in a fuzzy relation. The objective methods are performed in an experimental approach. The second class is a common one and assumes fuzzy model identification without an expert. This presumes an estimation algorithm for the fuzzy relation of a system based on an input-output data set. The paper presents an approach for identification of fuzzy relation based on square prediction error minimization and the results achieved through simulation on the Box-Jenkins data set