Advance 6-DOF Manipulator Controller Design Using DMRAC Based ANFIS

Abstract

Recent advances in artificial intelligent controller seems to be a more attractive to provide the fundamental driving force in many applications especially in areas of real-time manipulator control. In this paper, 6-DOF elbow manipulator controller using Direct Model Reference Adaptive Control (DMRAC) based hybrid Adaptive Neuro-Fuzzy Inference System (ANFIS) controller was proposed. Firstly, the theory of spherical-wristed 6-DOF mathematical modeling including the kinematic, dynamic solutions, different forces using different Jacobins is briefly formulated. Secondly, the ANFIS controller was hybrid with a classical controller to minimization the manipulator’s trajectory overshoot and the effect of settling time movement. Thirdly, the Virtual Reality Modeling Language (VRML) used in the whole manipulator-controller system and 3D display model simulated with MATLAB software Toolbox. Finally, based on performance of proposed method the simulation and on-line testing results show that the hybrid controller’s capacity of both precision and speed are improved compared with classical controllers.