Abstarct: To exploit the beneficial features of feedback linearization control and fuzzy control, and also to defeat their disadvantages, this paper presents an optimal robust adaptive fuzzy control scheme combining conventional feedback linearization control and an optimized compensator for the best robust tracking control of robotic manipulators with uncertainties in forms of structured and unstructured. The proposed compensator is baxsed on adaptive fuzzy estimation and compensation of uncertainty. This proposed adaptive fuzzy system is optimized aided particle swarm optimization algorithm and it can compensate the uncertainties by modeling of the uncertainties as a nonlinear function of the joint position variables. In addition the control system is adjusted such that the control signals are bounded. The advantage of the proposed adaptive fuzzy system is that tuned in best set of parameters, does not use all system states for estimating the uncertainty and the torque control signal will not exceed the saturation range. According to Lyapunov stability theory, a tracking error limit is derived for the closed‐loop control system and baxsed on it the convergence and stability of the control scheme is proved. In this thesis, after exxpression for the introduction of CNC machines, robot kinematic and dynamic modeling are presented. And then, feedback linearization control approach and after that, to deal with the parametric uncertainties adaptive control, and continue to work fuzzy-adaptive control are examined. In the end, the parameters of adaptive fuzzy controller optimized by particle swarm optimization. And through selecting the appropriate objective function, even with access to the minimum error, is prevented the torque signal controls to go beyond saturation range.