Neural network system for inverse kinematics problem in 3 DOF robotics

Abstract

Inverse kinematics computation has been one of the main problems in robotics research. An inverse kinematic analysis addresses the problem of computing the sequence of joint motion from the Cartesian motion of an interested member, most often the end effector. Traditional methods such as geometric, iterative and algebraic are inadequate if the joint structure of the manipulator is more complex. In addition, periodic characteristic of trigonometric resulted non-convexity of IKM. As alternative approaches, neural networks have been widely used for inverse kinematics modeling and control in robotics. The idea is to build a network that learned all the trajectory path of a model in different setting. Computer simulations conducted on 3DOF robot manipulator shows the effectiveness of the approach.