Feed Forward Predictive and Adaptive Control by the Dynamic Matrix

The basic concepts of an optimal predictive control algorithm referred to a Dynamic Matrix Control (DMC) are described. The Dynamic Matrix model is related to traditional models like state space and ARMA models. The control algorithm (DMC) is related to traditional optimal control in the unconstrained case. It is shown how constraints in input and output variables are defined and incorporated in the optimization. A brief description of an estimation scheme for the dynamic matrix parameters is given. Finally, an extension into nonlinear models and control is presented.