Trajectory tracking with obstacle avoidance of redundant manipulator based on fuzzy inference systems

In presence of the uncertainties in robot dynamics, model-based control law can easily fail. In our paper, this important question is tackled by using fuzzy adaptive control to drive with obstacle avoidance an industrial redundant manipulator under the hypothesis of uncertain dynamics. This hypothesis makes our controller an original one, since the problem of the tracking trajectory in the presence of obstacles and assuming that the robot dynamics are unknown has not been discussed before in the literature. Besides, the proposed obstacle avoidance is achieved by generating a self-motion. Furthermore, the usual procedure related to the obstacle avoidance based on Euclidean distance in 3D space is made easier since this one includes only the parameters of 2D space. This self-motion is directly incorporated into the adaptive fuzzy control scheme via the filtered tracking errors. The proposed method is tested by simulations in the case of redundant robot (PUMA 560) operating in 3D space in presence of obstacles. Despite that the robot dynamics are assumed to be uncertain, the proposed control performance is satisfactory. Indeed, the trajectory tracking control, incorporating self-motion obstacle avoidance, operates effectively with weak tracking control errors and bounded actuator torques evolving in admissible range.