Torque Estimation-Based Decentralized Control of Modular Robot Manipulators with the External Collisions

This paper proposes a decentralized control method for modular robot manipulators (MRMs) with external collisions based on torque estimation. Unlike some of conventional methods which depend on joint torque sensing technique, we address the torque estimation and motion control problems of MRMs based on only position measurements of each joint module. The dynamics model of MRMs is established by using the HD model and the estimated joint torque. By using only local dynamic information of each joint, a decentralized robust RBF neural network (NN) controller is designed to deal with external uncertain collision effects and joint trajectory tracking issues. The closed-loop system is verified to be asymptotic stable based on the Lyapunov theory. Finally, experiments are carried out to prove the effectiveness of the proposed method.