Abstarct: The discrete linear quadratic control has been efficiently used as an optimal control in linear certain systems. There would seem to be some difficulties to apply the discrete linear quadratic control to a robotic system which is highly nonlinear, heavily coupled and relatively uncertain. To overcome the problems, this thesis presents a novel optimal repetitive control of electrically driven robots using voltage control strategy. To apply the repetitive optimal control, a control-oriented discrete linear model for the robotic system with model uncertainty is introduced. The nonlinearities and uncertainties of the robotic system are compensated by the robust time-delay controller and the discrete linear quadratic control is performed. The control approaches are proposed in both torque control strategy and voltage control strategy. The torque control strategy is a commonly used strategy in the robotic system whereas voltage control strategy is recently developed. The robust optimal discrete repetitive control is free from manipulator dynamics using the voltage control strategy. Compared with the torque control strategy, it provides a better tracking performance. Then, the particle swarm optimization is used to design the optimal control and the obtained optimal control is compared with the proposed optimal control. The simulation results show that the proposed optimal control has a superior tracking performance to the particle swarm optimization control in the presence of uncertainties.