Model Predictive Tracking Control for Three-wheeled Mobile Robots Combined with Gaussian Process Regression

This paper studies the tracking control problem of the three-wheeled omnidirectional mobile robot (TOMR) subject to the coupled input constraint and unknown disturbances. First, a new type of coupled input constraint of three-wheeled omnidirectional mobile robots at kinematic level is derived. Second, we apply Gaussian Process Regression (GPR) on experimental data to compensate for the disturbances. Third, the trajectory tracking control algorithm is proposed based on nonlinear model predictive control (NMPC) method and Gaussian Process Regression. To reduce the computational burden of the control law, three efficient improved approaches are proposed. Finally, we generalize the designed control strategy to solve the popular collision avoidance problem. Numerous simulations are carried out to validate the theoretical results.