Optimization of DC Microgrid Power and Energy Management in the Presence of Small-Scale Wind Turbine Integration and Renewable Load Shedding

Abstract

Improvement of Direct Current Microgrid (DCM) Power and Energy Management is the methodical process of increasing the DCM system’s performance, dependability, and effectiveness. In this study, the suggested a Stochastic Dove Swarm Optimized Deep Reinforcement Learning Algorithm approach to estimate the availability of wind energy, allowing for proactive management tactics. Initially, the data was collected and preprocessed by performing min–max normalization to clean the unwanted data. Recursive Feature Elimination is used to determine the greatest informative attributes. This study analyzes the use of significant performance metrics, including accuracy (91%), precision (92%), recall (93%), Root Mean Square Error (0.154), Normalized Root Mean Square Error (0.0526), and Mean Absolute Percentage Error (4.21). The use of optimization methodology focuses on the seamless integration of SWTs and the application of renewable LS methods to ensure the efficient management of power and energy in DC microgrids. The intention is to keep operational expenses to a minimum while providing a stable and robust energy supply. To ensure power balance, which serves as the primary objective of the monitoring, it is considered imperative for carrying out electricity limitations in this instance of the energy excess generated by solar power plants and the small-scale wind turbine.