Abstarct: This thesis is concerned with the robust nonlinear control of electrically driven robot manipulators and presents a novel method baxsed on voltage control strategy to compensate the tracking error caused by uncertainties including parametric uncertainty, external disturbances and un-modeled dynamics. The voltage control strategy is superior than the conventional torque control strategy since it is free of manipulator dynamics. Consequently, the computational load of the controller is less in comparison with the torque control strategy. The most important novelty of this thesis is using the particle swarm optimization algorithm (PSO) to minimize the tracking error in the joint-space. The stability of the system is guaranteed through the Lyapanov direct method and the PSO is employed to enhance the control system performance. The case study is a spherical robot manipulator driven by permanent magnet dc motors. Simulation results illustrate the capability of the PSO algorithm in reducing the tracking error. Moreover, the comparison of the voltage control strategy and the torque control strategy reveals the superiority of the voltage control strategy.