Abstarct: In this thesis, an adaptive sliding mode control scheme for a robot manipulator baxsed on voltage control strategy is presented and the proposed method were compared with adaptive inverse dynamics. Adaptive controllers in presence of unstructured uncertainties such as disturbance and unmolded dynamics do not perform properly but these controllers can overcome structured uncertainties truly also sliding mode controllers can overcome unstructured uncertainties but the weakness of these controllers is specifications of uncertainty boundaries. If uncertainty boundaries is chosen smaller than an accurate amount then tracking error will be increased otherwise if uncertainty boundaries is chosen greater than an accurate amount then it causes input saturation and chattering of control command. The proposed method in this thesis can overcome structured uncertainties and unstructured uncertainties such as unmolded dynamics and external disturbance easily. In this method there is no need to calculation of the inverse of Inertia matrix and measurements of joints acceleration and since torque command cannot be applied directly to robot so the dynamics of engines and gears are added to the robot. This act makes production of voltage instead of torque as the input control of the system be more straightforward. Using of gear reduction decreases complex couplings of manipulators and its control scheme makes stability be more reliable. At the end the stability of the proposed method is proved and its performance compared with the adaptive inverse dynamics is evaluated which the results show the superiority of the proposed method.