Abstarct: In this thesis, several model-Free control methods have been used for robot manipulators control. The general structure of manipulator in this thesis is baxsed on Voltage Control Strategy. In this structure, the unknown parameters of the system, such as disturbances and uncertainties, are considered and as a single variable. In the proposed methods, it is not necessary to complex robot model for control. Each joint is controlled independently and the control signal generated by the control mechanism is the voltage of the motors, which can be applied to them directly. The implementation of these methods are easy. The presented methods in this thesis are Model-Free Control using an extended state observer, Adaptive Model_Free Control and Model_Free Control using Neural Network. The results of mathematical analysis and simulations and comparisons show that these methods are effective.