Obstacle avoidance for kinematically redundant robots using an adaptive fuzzy logic algorithm

In this paper the adaptive fuzzy logic approach for solving the inverse kinematics of redundant robots in an environment with obstacles is presented. The obstacles are modeled as convex bodies. A fuzzy rule base that is updated via an adaptive law is used to solve the inverse kinematic problem. Additional rules have been introduced to take care of the obstacles avoidance problem. The proposed method has advantages such as high accuracy, simplicity of computations and generality for all redundant robots. Simulation results illustrate much better tracking performance than the dynamic base solution for a given trajectory in Cartesian space, while guaranteeing a collision-free trajectory and observation of a mechanical joint limit.