Design of a Load Frequency Controller Based on an Optimal Neural Network

A load frequency controller (LFC) is a crucial part in the distribution of power system network (PSN) to restore its response when demand changed rapidly. In this paper, an artificial neural (ANN) technique utilised design optimal LFC. However, training ANN model for multi-area PSN major challenge due variations demand. To address challenge, particle swarm optimization used distribute nodes hidden layer and optimise initial neurons model, resulting obtaining lower mean square error model. Hence, number epochs are minimised about 9.3886 × 10−8 25, respectively. assess proposal, MATLAB/Simulink developed single-area PSN. The results show that LFC based on more effective adjusting level improves delivery comparison with Moreover, it most reliable avoiding fault condition whilst achieving lowest time multiplied absolute 3.45 s compared conventional PID methods.