Generating crisp-type fuzzy models from operating data

A simple method for generating a multiple-input/single-output crisp-type fuzzy model from observed data is presented. The crisp-type fuzzy model is composed of triangular fuzzy partitions for inputs and singletons for the output. An explicit input-output equation of the crisp-type fuzzy model is supplied. For a set of observed input/output data pairs, the structure and the parameters of a suitable fuzzy model with required accuracy can be determined by a series of algebraic computations. Numerical examples are given to demonstrate the effectiveness of the proposed identification method for modeling static and dynamic data.