Velocity-Free Prescribed Performance Control of Dual-Arm Manipulator Based on Neural Networks

The prescribed time control of dual-arm manipulator transferring a target without velocity measurement is discussed in this brief. Firstly, a comprehensive dynamic model is formulated using the target position in Cartesian space. Secondly, the prescribed-time tracking problem is simplified to an error constraint problem, where a time regulator is constructed, and a constraint transformation function is established. Thirdly, algorithms for position control are proposed based on radial basis function neural networks (RBFNN), along with the development of comprehensive position/force control, followed by proof of their effectiveness. Finally, the validity of these results is demonstrated through numerical simulations.