In this paper, a novel control strategy for the load frequency control (LFC) system is proposed. The developed method includes a genetic algorithm (GA) based self-tuned PID controller for online application. The new method is presented in order to regulate PID controller coefficients by a radial basis function neural network (RBFN). Furthermore, a very short time load forecasting (VSTLF) scheme is also employed as a novel approach for the system load variations to be considered in the LFC system. For validation of the proposed method, several comparative case studies are presented. The simulation results indicate that the proposed strategy improves the system dynamics remarkably.