Neural Network Based Global Adaptive Dynamic Surface Tracking Control for Robot Manipulators

Tao Teng,Chenguang Yang,Bin Xu,Zhijun Li
DOI: https://doi.org/10.1109/icarm.2016.7606888
2016-01-01
Abstract:A neural network empowered dynamic surface control (DSC) technique is addressed for robot manipulators system with unknown dynamics. In comparison to the conventional adaptive neural control algorithms, which could guarantee semi-globally uniformly ultimate boundedness (SGUUB) only when neural approximation keeps effective, the scheme designed in this paper ensures globally uniformly ultimately bounded (GUUB) stability by integrating a switching mechanism which incorporates an additional robust controller to drag the transient state variables back when they go beyond the neural approximation region. Simulation studies on 2-joint robot manipulator have been carried out to validate the designed controller has excellent performance.
What problem does this paper attempt to address?