Nonlinear model-predictive control of Wiener-type systems in continuous-time domain using a fuzzy-system function approximation

摘要

This paper presents a method of continuous-time model predictive control for nonlinear Wiener-type systems. The system's output nonlinear mapping is approximated by the means of fuzzy systems, and is inherently incorporated in the predictive control scheme; thus, inverting the output function is not necessary. The predictive control law is derived in an analytical form to avoid the problems with non-convex optimization in each time step. A simulation experiment is conducted to present the control quality in the case of a pH-neutralization process with a heavily nonlinear output mapping.