A Collaboration Scheme for Controlling Multimanipulator System: A Game-Theoretic Perspective
Jiazheng Zhang, Long Jin, Yang Wang, Chenguang Yang
Abstract
In some task-oriented multimanipulator applications, the system not only needs to complete the main assigned tasks, but also should optimize some subobjectives. In order to tap the redundancy potential of individual manipulators and improve the performance of the system, a hybrid multiobjective optimization solution with robustness is proposed in accordance with the realistic execution requirements of the tasks. The entire control scheme is designed from the perspective of the Nash game and further refined into a problem to determine the Nash equilibrium point. Furthermore, a neural-network-assisted model is established to seek the best response of each manipulator to others. Theoretical analysis provides support for proving the convergence and robustness of the model. Finally, the feasibility of the control design is illustrated by simulation studies of the multimanipulator system.