Output end-point weighted generalized predictive control-a polynomial approach

A new generalized predictive control (GPC) algorithm is presented, which makes use of the concept of output end-point weightings. The aim of the algorithm is a reduced computational burden with respect to the state end-point weighted GPC and a relaxation of the infinite output weighting of constrained receding horizon predictive control. The new output end-point weighted GPC algorithm is developed using the polynomial approach. After a proof of nominal stability the relation of Kalman filtering and predictive control is derived. A final simulation of a benchmark example emphasizes the reduced computational burden and the enhanced stochastic tracking capability