Model-based path planning and tracking control using neural networks for a robot manipulator

In this paper, a new method is presented for a path planning and tracking control of a planar robot manipulator in the presence of system uncertainty and obstacles using neural networks combined with the conventional feedback controller. Our path planning provides not only the (sub)optimal trajectory for a given cost function through evolutionary algorithm but also the configurations of the robot manipulator along the path by considering the robot dynamics. The path can be made to keep away from the singular points and avoid the obstacles. An additional neural controller compensates for the tracking errors caused by uncertainty and disturbance, which provides the robustness with a good tracking performance. Computer simulations show the effectiveness of the proposed method