In this paper, a hybrid fuzzy controller with a real-time particle swarm optimization algorithm (PSO) is presented to swing up and to stabilize the pendulum-cart system. The designed PSO with a re-start mechanism is particularly suitable for dynamic environments and its feasibility is evaluated using the typical dynamic functions. The modified particle swarm optimization is performed to on-line tune the parameters of hybrid fuzzy controllers. Also, the stability of the control system is discussed for the PSO optimized stable fuzzy control system. Simulation results, compared to conventional fuzzy control, illustrate that the proposed optimization-based control scheme can provide better control performance subject to extra disturbances.