Finite-Time Neural Impedance Control for an Uncertain Robotic Manipulator

This paper proposes a finite-time neural impedance control for a robotic manipulator. A position-based impedance controller is proposed to improve the safety and compliance when robotic manipulator contacts with environment physically. Radial basis functions neural networks (RBFNNs) are employed to compensate uncertainties in robotic manipulator dynamics. A finite-time control method is developed with the back-stepping technique to improve the tracking performance. Large external forces can be avoided and desired impedance model can be achieved quickly under our proposed method. The stability in the close-loop system is proven by Lyapunov theory, and all error signals in the system are semi-global practical finite time stable (SGPFS) and the system output converges to reference signals in finite time under our proposed controller. Finally, comparative simulations are proposed to verify the effectiveness of our proposed method.