A comparative study of PID and neuro-fuzzy based control schemes for a 6-DoF robotic arm

Abstract

The robotic arm is one of the most prominent manipulators used in the field of robotics and therefore serves as a benchmark problem in control systems. Recent developments in the field of artificial intelligence demonstrate the use of intelligent control techniques for various systems. This paper presents a neuro-fuzzy based control scheme for a robotic arm with 6 degrees of freedom used for the sorting and placing of coloured balls. A webcam based image acquisition scheme is used to locate and identify various colored balls and subsequently the robotic arm is moved according to required trajectory. The performance of the proposed neuro-fuzzy controller is compared with a traditional PID control to illustrate the superiority of the intelligent control scheme. Experiments performed on a hardware test bench validate the improvement in performance achieved using the neuro-fuzzy control scheme.