It is difficult to control for the level of the steam generator in the nuclear power plant because there is swell and shrink, and many disturbance such as, feed water rate, feedwater temperature, main steam flow rte, coolant temperature effect steam generator level. If the conventional PI controller use in this system, we cannot have a stability in the control of the lower power, the rejection function of disturbance, and the load following effectively. In this paper, e study the application of the of neural network based Kp, Ti for Pi controller to the level control of the steam generator of nuclear power plant through the simulation and experimental on the steam generator. We are satisfied with the resulting against the inturrupt of the disturbance, the change of setpoint through the simulation and the swell and shrink, the response of controller on the experimental steam generator.