Decentralized control with neural network of cooperative robot manipulator for object balancing task on flat plate

In this paper, a decentralized framework for kinematic control of cooperative manipulators systems is developed. The motion of the robot system is specified at the object position, by adopting a task-oriented formulation for cooperative tasks. In the controller of robot computes, the end-effector motion of robots in a decentralized on the camera position frame of the knowledge of the assigned cooperative task. The motion of manipulator is reference computed by object and its neighbors. The joint motion of robots is reference corresponding from the end-effector. This study approach decentralized control of collaborative manipulators, tested in the simulation on MATLAB^® software composed by neural network method. A neural network is used to approximate a decentralized control law designed by the back-propagation technique. The motion for each joint is controlled independently using local angular position and velocity measurements. Finally, the experiment shows the feasibility of the proposed control scheme using a robotics.