Optimal Control for Robotic Manipulators With Input Saturation Using Single Critic Network
Guangran Cheng, Lu Dong · 2019
This paper proposes an optimal control scheme for complex robotic manipulators with input saturation and dynamical disturbances. A single critic neural network(NN) is adopted in the controller design. In addition, a suitable nonquadratic function is provided to tackle the problems of constrained inputs and system uncertainties. Furthermore, an updating law for critic NN is used to guarantee that the closed-loop robotic system is ultimately bounded. The stability analysis is also addressed based on the direct Lyapunov method. Finally, the simulation is conducted on a two degree-of-freedom(DOF) robotic manipulator to verify the effectiveness of the proposed optimal control scheme.