Distributed Collaboration and Obstacle Avoidance Control of Multiple Omnidirectional Mobile Manipulator Assisted with Neural Networks

In this paper, a new distributed control scheme is proposed to realize the cooperative motion and obstacle avoidance control of multiple omnidirectional mobile manipulator (OMM) under the premise of limited communication. The proposed scheme enables multiple OMM to avoid end-effector squeezing phenomenon during cooperative load transportation, and to avoid point obstacles by introducing a cooperative repetitive motion planning and obstacle avoidance mechanism. Then, the proposed scheme ensures global convergence to the optimal solution by solving online utilizing noise suppression neural network (NSNN). Numerical simulation results verify the effectiveness of the control scheme to realize cooperative obstacle avoidance and efficient load transportation in complex environments.