Superiority of Nonlinear and Stable MPC in a Differential Mobile Robot: Accuracy and Solving Speed

nowadays, the need to robots is increasing daily in industry and science. For that reason, researchers are trying for developing of accuracy for the function of robots in all applications. In some applications not only accuracy is important, but also solving speed is. In this paper, these two points are followed for control of a differential mobile robot. The stable model predictive control (SMPC) is used for robot path tracking by using the kinematics equations of the robot to follow the first point, accuracy. For this controller, the kinematics equations are linearized by the tracking error method and they are used. Also, we have used the nonlinear model predictive control (NMPC) to follow both of the first and second points in the same time. This point is important that when the equations are linearized we are making more errors for robot control. Although, using the SMPC was good to increase accuracy, but we had to use a function to solve the extra equation that SMPC has. The problem of this method is the long time of the solving. While NMPC has more process in every step for solving equations in compare of model predictive control (MPC), but the time of the solving it is less than the time of SMPC solving. If the processing and solving speed is low, the robot will pass the desired step and SMPC will become practically meaningless. This speed of the NMPC is higher about 3 times than the SMPC, which is better performance in this view as a controller.