Adaptive Neuro-fuzzy Inference System Based Control of puma 600 Robot Manipulator

The strong dependence of the computed torque control of dynamic model of the robot manipulator makes this one very sensitive to uncertainties of modelling and to the external disturbances. In general, the vector of Coriolis torque, centrifugal and gravity is very complicated, consequently, very difficult to modelled. Fuzzy Logic Controller can very well describe the desired system behavior with simple “if-then” relations owing the designer to derive “if-then” rules manually by trial and error. On the other hand, Neural Networks perform function approximation of a system but cannot interpret the solution obtained neither check if its solution is plausible. The two approaches are complementary. Combining them, Neural Networks will allow learning capability while Fuzzy-Logic will bring knowledge representation (Neuro-Fuzzy). This paper presents the control of puma 600 robot arm using Adaptive Neuro Fuzzy Inference System (ANFIS) based computed torque controller (type PD). Numerical simulation using the dynamic model of puma 600 robot arm shows the effectiveness of the approach in improving the computed torque method. Comparative evaluation with Fuzzy computed torque (type PD) control is presented to validate the controller design. The results presented emphasize that a satisfactory trajectory tracking precision and stabilility could be achieved using ANFIS controller than Fuzzy controller. Keywords: Fuzzy computed torque control, Robot control, Adaptive neuro-fuzzy inference system (ANFIS).