An Adaptive Backstepping-Based Controller for Trajectory Tracking of Wheeled Robots

With recent development of technologies in mobile robotics, wheeled robots are playing a more and more important role in unmanned transportation and exploration. One of the remained challenges in this research area is trajectory tracking control that aims at improving tracking accuracy and ensuring stability of non-holonomic constrained dynamic systems. Based on the mathematical model of wheeled robot system, this paper proposed a backstepping-based controller that self-tunes its control parameters by a neural network, so as to enhance the adaptability of the algorithm. At the same time, at the present stage, people from all walks of life are widely concerned about energy consumption. Saving energy means reducing consumption and protecting resources and environment. Mobile robots commonly carry batteries as their power source, and the operating time is limited by the remaining power of the batteries. Therefore, in this work we optimize energy consumption of the robot as an index in the neural network. The effectiveness of the algorithm is verified by simulation, and primary results show that the proposed controller can achieve accurate trajectory tracking while minimizing energy consumption of the wheeled robot.