Model predictive controller design

Model Predictive Control (MPC) is a very powerful control strategy that manipulates multivariable constrained systems. It is quite different from classical control and requires high design capabilities. The main aim of this thesis is to apply one of the MPC strategies to control a DC motor by simulations. For this, first of all, a mathematical model of the DC motor based on real parameter values of a DC motor has been derived. Using derived model of the DC motor, the motor has been controlled based on one of the MPC strategies. The effect of tuning parameters of MPC on the control performance has also been investigated. In order to test the performance of designed MPC, studies to control a DC motor based on classical control Proportional-Integral-Derivative (PID) has also been carried out. A cascade control scheme to control both the current and angular velocity of the DC motor has been used. In this scheme, the PI controller in the inner loop was designed to control the current of DC motor and PI-P controller in the outer loop was designed to control angular velocity of the DC motor. Tuning parameters of controllers' both in the inner and outer loop was determined using the controller synthesis approach. Closed loop performances obtained from the cascade control structure and liner MPC designed in MATLAB/Simulink environment and MPC Toolbox have been compared. Finally, the closed loop performance of the designed MPC and cascade control structure under different working conditions has been investigated.