Automatic Load Frequency Control for Wind-Thermal Micro Grid Based on Deep Reinforcement Learning

Abstract

Renewable energy demand keeps increasing each day due its significances over the conventional sources of energy, particularly in this era where the world is faced with many challenges related to clean energy. Among Renewable Energy Resources (RERs), wind energy has proven to be cheaper and readily available. However, it is intermittent in nature and therefore affecting the voltage and frequency stability of microgrid systems, especially in occurrence of wind power ramping events. In this work, a simple Deep Reinforcement based Automatic Load Frequency Controller (DRL-ALFC) is designed so as to improve the frequency stability of an ALFC during wind power ramping events in a wind-thermal micro grid. A DRL-ALFC for wind-thermal microgrid is verified in MATLAB/Simulink environment where it shows the ability to adapt to the variations wind power fluctuation and load.