Design and Evaluation of a MIMO ANFIS using MATLAB and V-REP

Employing software computing tools, such as fuzzy logic, neural networks and genetic algorithms, encouraged researchers to design intelligent systems that mimic human behavior. The Adaptive Fuzzy Nervous System (ANFIS) is one of the most useful hybrid technologies for designing intelligent control algorithms. Designing a Multiple-Input, Multiple-Output (MIMO) ANFIS algorithm, is not an easy task for those with limited programming experience. This paper presents a simple way to design MATLAB-based MIMO ANFIS controllers for two mechatronics plants; a two-wheeled mobile robot and a 6DOF KUKA robotic arm. The 3D simulation software (V-REP) has been interfaced directly with the MATLAB Simulink to test the proposed controller. The generated 3D simulation environment given in this paper will help researchers to easily design and test of MIMO ANFIS algorithms. The obtained results demonstrate the performance of the mobile robot and ability of the implemented ANFIS controller to deal with input commands, obstacle avoidances in the working environment.