Analytic approach to generaised predictive control of nonlinear systems

An analytical approach to generalized predictive control (GPC) of nonlinear systems was discussed. It was found that almost all engineering systems were described by continuous-time differential equations and it was natural to develop GPC in a continuous-time setting. The results were based on four concepts mainly prediction via Taylor series expansion, receding horizon control, control constraints and optimization. The features of this new nonlinear predictive control law were: no on-line optimization was necessary, the design parameters and the predictive time could be determined and the stability of the closed loop system was guaranteed for the nonlinear systems.