A New Heuristic Approach for Inverse Kinematics of Robot Arms

Abstract

There have been many topics related to robotics, such as motion planning, intelligent control and tracking control. The changes in joint space generate a motion in Cartesian space. The set of joint variables should be determined to obtain a desired position of the end-effector in Cartesian coordinate. Thus, there is a need to translate from Cartesian space to joint space, which is known as inverse kinematics, while forward kinematics deals with transformation from joint space to Cartesian space. Many algebraic 1,2,3,4 and numerical 5,6 methods have been developed for inverse kinematics. Inverse kinematics problem is suitable for the heuristic methods because it has multiple solutions. These methods produce better solutions even for the NP problems in a reasonable time 7 . In the recent years, bio-inspired algorithms have been applied for inverse kinematics problem for five or more degree-of-freedom (DOF) robot arms because of offering more accurate results than the traditional methods. Neural Networks 8,9 , Genetic Algorithm (GA) 10 , Harmony Search (HS) 11 and Particle Swarm Optimization (PSO) 12