Lyapunov function based neural networks for adaptive tracking of robotic arm

In this paper, we aim to present an adaptive position controller for multiple degree of freedom robotic manipulators. A decentralized approach is presented that utilizes Lyapunov function based artificial neural networks as inverse controllers of the robot’s nonlinear coupled dynamics. The proposed scheme is successfully implemented on the real time control of the TQ MA3000 robotic manipulator. Promising experimental results show the effectiveness of the proposed algorithm in the sense of fast convergence of adaptive tracking error and stability of the closed loop.