Control-Oriented Fuzzy Multi-Model Identification of a Highly Nonlinear Missile

Abstract

A novel fuzzy identification algorithm is introduced that is based on fuzzy interpolation of locally linear models obtained from a nonlinear system. Fuzzy clustering is employed to automatically decompose operating envelope of a nonlinear system into optimal number of operating partitions by means of optimal number of rules in a Tagaki-Sugeno fuzzy set direct realistic extension to flight control is introduced and applied on a tail-controlled skid-to-turn missile. The fuzzy model is evaluated using simulations of the complete nonlinear model. Simulation studies are reported to demonstrate merits of fuzzy systems in handling the demanding nonlinear modeling task