Abstract: In this paper, online policy iteration reinforcement learning (RL) algorithm is proposed for motion control of four wheeled omni-directional robots. The algorithm solves the linear quadratic tracking (LQT) problem in an ...