Object Manipulation of Humanoid Robot Based on Combined Optimization Approach

Abstract

In this paper an object manipulation approach is proposed, in which object location is detected through stereo vision, pre-reaching and alignment of the object with hand is based on image features, and finally inverse kinematics control algorithm is developed. A new method for computing numerical solution to the inverse kinematics problem of robot manipulator is developed. The proposed method is based on combination of two nonlinear programming techniques. Forward recursion formula with backward cycle computation method is used to reach in the vicinity of target object and then FBS method is used to grasp object in real-time. Proposed method is numerically stable, computationally effective and it is not sensitive to singular configuration of the manipulator. Effectiveness of the proposed method is achieved through simulated and experimental results.