Trajectory optimization and tracking control of mobile robots based on DMOC

Abstract

This paper is based on a new trajectory optimization approach -Discrete Mechanics and Optimal Control (DMOC). The optimal trajectory generation and trajectory tracking controller of mobile robots are investigated. Firstly, DMOC is applied to generate an optimal reference trajectory which is constrained by specific optimization conditions to meet the given cost functional. Secondly, considering the high nonlinear and uncertainties are inevitably present in mobile robots, the reference trajectory is utilized in the proposed adaptive trajectory tracking controller with Backstepping concepts. The control system is on the basis of the dynamic equations and the real-time velocity vectors are served as control inputs. Finally, with the application of the optimal trajectory simulation platform, the control effect is verified. The simulation results prove that DMOC approach has certain accuracy, universality and applicability, besides the high nonlinear, strong coupling system with nonholonomic constraints. The designed trajectory tracking controller has good robustness and efficiency which is capable of adjusting the unknown parameters in real-time and realizing accurate and stable tracking control.