Model predictive control of nonholonomic mobile robots

Abstract

In this work, we investigate the possibility of using model predictive control (MPC) for the motion coordination of nonholonomic mobile robots. The contributions of this dissertation can be summarized as follows. A robust formation controller is developed for the leader-following formation of unmanned aerial vehicles (UAVs). With the assumption that an autopilot operating in holding mode at the low-layer, we present a two-layered hierarchical control scheme which allows a team of UAVs to perform complex navigation tasks under limited inter-vehicle communication. Specifically, the robust control law eliminates the requirement of leader's velocity and acceleration information, which reduces the communication overhead. A dual-mode MPC algorithm that allows a team of mobile robots to navigate in formations is developed. The stability of the formation is guaranteed by constraining the terminal state to a terminal region and switching to a stabilizing terminal controller at the boundary of the terminal region. With this dual-mode MPC implementation, stability is achieved while feasibility is relaxed. A first-state contractive model predictive control (FSC-MPC) algorithm is developed for the trajectory tracking and point stabilization problems of nonholonomic mobile robots. The stability of the proposed MPC scheme is guaranteed by adding a first-state contractive constraint and the controller is exponentially stable. The convergence is faster and no terminal region calculation is required. Tracking a trajectory moving backward is no longer a problem under this MPC controller. Moreover, the proposed MPC controller has simultaneous tracking and point stabilization capability. Simulation results are presented to verify the validity of the proposed control algorithms and demonstrate the performance of the proposed controllers.