Stabilization of System Frequency using Adaptive Load Shedding Scheme Based on Artificial Neural Network

Safe operation of a system requires the voltage and frequency within safe limits. Recent trends show that large disturbance are more frequent and directly affects the frequency stability. Load Shedding is the last control action that can be taken to prevent complete blackout of the system. In this paper, frequency is stabilized by Artificial Neural Network (ANN) based load shedding. The ANN is trained by Levenberg-Marquardt back propagation method, with power mismatch and generator turbine constants as the input neurons of the ANN in MATLAB. The effectiveness of the scheme is verified in a 6 bus IEEE system developed in Power World Simulator.