A novel approach to develop the control of Telbot using ANFIS for nuclear hotcells

Abstract This paper develops an improved controller for Telbot. Telbot is a new 6 DOF tele-robot, with special kinematic design, famous for working in nuclear hotcells. The recent controllers are based on simple and improved PID controllers, with visible tracking error. In this project, an ANFIS controller was designed and trained for controlling this tele-robot. The proposed controller generates an appropriate torque for reaching desired state, without any error. The controller was initialized with if-then rules. The training includes online and offline processes. The offline processes have light loads, and prepaid more than 90% of system desired, but the online processes do the last critical 10%, and remove all the errors. The controller is describing the high degree of nonlinear dynamics behavior of the Telbot. The final simulation shows the error-less tracking.

[1]  Friedrich Pfeiffer,et al.  Control of a new tele robot , 2000, 6th International Workshop on Advanced Motion Control. Proceedings (Cat. No.00TH8494).

[2]  Frank L. Lewis,et al.  Neural Network Control Of Robot Manipulators And Non-Linear Systems , 1998 .

[3]  Jyh-Shing Roger Jang,et al.  ANFIS: adaptive-network-based fuzzy inference system , 1993, IEEE Trans. Syst. Man Cybern..

[4]  Chuen-Tsai Sun,et al.  Neuro-fuzzy modeling and control , 1995, Proc. IEEE.