LENGTH-OPTIMIZED SMOOTH OBSTACLE AVOIDANCE FOR ROBOTIC MANIPULATORS

Abstract

This paper presents a novel optimized smooth obstacle avoidance algorithm for robotic manipulators. First, a 3-4-5 interpolating polynomial is used to plan a smooth trajectory between initial and final positions in the joint space without considering any obstacles. Then, a simple harmonic function, which is smooth and continuous in displacement, velocity and acceleration, is applied to generate a new smooth path avoiding collisions between the robot links and an obstacle. The obstacle avoidance portions on the path are optimized such that the length of the path traversed by the end-effector is minimized. Simulation results for a 6 DOF serial manipulator demonstrate the efficiency of the proposed method.