A Simplified Dynamic Matrix Control for application of embedded real-time MPC

Model predictive control (MPC) has been proved to be an effective advanced control method in the petrochemical industry and extended to other areas gradually. Dynamic Matrix Control (DMC) is an early widely used representative algorithm of MPC. As quick and powerful computing requirements of normal DMC, it is difficult to apply it to the real-time control with a small-physical-size embedded system. In order to extend the MPC application, in this paper, a shifted matrix is introduced to make the control increment convergence to zero in the control horizon. Computing a number´s reciprocal can substitute for the matrix inversion and the control horizon need not be one in the simplified algorithm. The simulation confirms the validity of the algorithm. The problem with a simple constraint can be solved by modifying parameters of the shifted matrices.