engleski

Combining Identification and Constrained Optimal Control of Piecewise Affine Systems

In this paper we focus on identification and time optimal control of nonlinear processes modeled as piecewise affine systems. We combine the piecewise ARX process model identification based on clustering and the constrained time optimal controller design for discrete-time piecewise affine systems. The two procedures are improved and bound into a systematic procedure for design of high-performance nonlinear control systems: from the identification data to the closed-form time optimal controller. We successfully experimentally verify the procedure on the electronic throttle control system case study.

piecewise affine model; model identification; clustering; constrained time optimal control; electronic throttle

nije evidentirano