NEURAL NETWORKS FOR LOAD-FREQUENCY CONTROL

This paper presents a neuro-controler design for load-frequency control of a single - area system. It controls the system admission valve of the turbine In order to restore the frequency to its nominal value while enhancing the transient behavior. The inputs of the neuro-controfier include the frequency variation error and change of error frequency variation and the amplitude of load perturbation. A reference model was used to find the desired traiectory of frequency variation. To show the ability of the proposed con~er some digital simulations have been performed on a typical singlesystem. The results show that the proposed method makes the overal system have satisfactory dynamic response over the wide range of load changes.