Neuro-fuzzy based sliding mode control of a four degree of freedom robot arm

Robot arms are being used in industrial applications such as material handling and continuous production lines. In particular, four joint robot arms have high maneuverability and high precision capability required in industry. This can be provided by using control systems, but effects of some disturbances such as external applied torques and high speed working conditions should be taken into consideration during designing of controllers. In this study, a neural fuzzy based sliding mode control (FNSMC) structure had been designed, and the performances of that were evaluated with simulation works.Neural network, fuzzy logic, sliding mode, and computed torque methods are the most common in the control of robots. However, the use of each method is not individually sufficient to provide the required performance in the complex kinematic and dynamic structured robots. The neural fuzzy sliding mode controller has been developed to track the given trajectories in the best way. The designed FNSMC with computed torque method (FNSMCCTC) was aimed to enhance the feedback control.The obtained results have shown that both controllers have similar properties, but the FNSMCCTC was more successful than the FNSMC in the presence of disturbances.Keywords: Four degrees of freedom robot arm, neuro-fuzzy based sliding mode control, computed torque method, redundant degree of freedom problem.