Impact of energy storage devices on microgrid frequency performance: A robust DQN based grade-2 fuzzy cascaded controller

Abstract

This article intended an advanced deep-Q network (A-DQN) based algorithm for the optimal design of a robust grade-2 fuzzy cascaded controller (G2-FCC) concerning frequency control of an autonomous AC microgrid under several electrical disturbances. The proposed AC microgrid is equipped with a hydraulic electrolyzer fuel cell (HE-FC) that empowers electrically to a variety of local consumers. Apart from the HE-FC, the proposed microgrid model is associated with few fractional megawatts (MW) rated power generating sources such as diesel generators (DEG), micro-turbines (MT), solar power stations (SPS), wind power stations (WPS) and geothermal to maximize the total power of a system. However, to increase system performance and energy quality some energy-storing devices such as ultra-capacitors (UC), flywheel (fw), and battery energy management systems (BEMS) are integrated into the system. Most renewable energy sources such as solar and wind power plants are associated with high uncertainties and exhibit low inertia which severely affects the grid frequency. The proposed fuzzy G2-FCC controller dominates the frequency disturbance gracefully and shows its superiority over a few standard approaches like fuzzy Proportional Integral Derivative (PID) and PID controllers. The energy-storing components also have shown their potential in regard to frequency profile improvement of the microgrid system. Finally, the suggested A-DQN algorithm is proved as the most effective tool to optimal design the parameters of the G2-FCC controller. Finally, it is verified that the suggested G2-FCC controller is found to be the most effective tool to improve the settling time of ΔF by 154.54 and 249.64 % over well-existing grade-1 fuzzy PID and PID approaches, respectively.