An on-line tunable neural network position regulator for single-link robotic arm

The paper presents the development of an online self tunable neural network position controller for a single link robotic arm. The use of this controller is motivated by the arm model uncertainty, parametric variations, and sudden load excursions. The objective of the controller is to make the robot arm position follow a specified reference trajectory with minimum steady state error, and no overshoot/undershoot. The laboratory experimental results indicate that the proposed online tunable neural network position controller is robust and effective regulator