An approach toward Human-like Motion Control of a Dual Arm Robot for Picking Heavy Objects

This paper presents a novel method for controlling a dual arm robot to pick heavy objects. The motivation for this study is inspired from the observation of human motion in their daily activities. Two reflex actions are proposed in this work. The first reflex is designed to control the COM of the robot when its stability is endangered by the effect of the object weight. Static stability criterion is used to observe this dangerous effect and a simple minimization technique is used to generate the reflex torque. The second reflex aims to reduce the consumed torque of the motors by simply finding a pose at which the gravity term minimized. A control law is suggested for a dual arm robot to perform picking motion for both two cases with known and unknown objects. The control state space is the operational space and no inverse kinematics are needed. Finally, several dynamic simulations are introduced to validate the efficiency of the proposed controller.