Abstract
Synchronised motion is an important requirement for two cooperating humanoid robot arms. In this work a cooperative scenario is considered where two humanoid robot arms (using 4DOF each, namely Shoulder Flexion Joint, Shoulder abduction Joint, Humeral rotation joint and Elbow Flexion Joint) motion are synchronized. The master robot arm is controlled by a sliding mode controller and the slave robot arm is synchronized using a basic PD plus adaptive control, employing the position and velocity errors between the master and the slave. During the operation, if a joint of the slave robot arm saturates or malfunctions (for instance, Elbow flexion joint does not respond or free swinging), consequently, slave robot arm will go into chaos (i.e., chaotic motion of the end effector). In this case, a chaos controller kicks in to recover and re-synchronize the motion of the slave robot arm end effector. This re-synchronization is extremely important to complete the task in hand to address any safety issues arising from any joint malfunction of the slave robot. Effectiveness of the scheme is tested in simulation using Bristol Robotics Laboratory Humanoid BERT II arms.