A simplified constrained multivariable controller with steady-state optimization

A simplified MIMO constrained controller is proposed. Based on the steady-state prediction of system outputs, the controller combines steady-state optimization and dynamic control calculations. A quadratic programming approach is employed to find an optimal feasible solution. The controller is applied to the Shell benchmark problem. Compared with DMC-LP/sup /spl reg// controller, the simulation results show that the new controller not only achieves good control performance, but also costs less in computation time.