Fuzzy identification of nonlinear systems

This paper presents a mathematical way of building a fuzzy model of any nonlinear system. The fuzzy implications of the system model and the least square identification method have been used to describe the nonlinear systems under study. The phase plane on which the nonlinear system is to be represented has been partitioned into fuzzy subregions and a linear fuzzy system model has been identified for each region. Then it has been observed that the overall system behavior has been characterized quite satisfactorily by using this partitioned fuzzy modelling.>