Bi-Objective Optimization for Intelligent Microgrids: A DRL Approach with Rat Swarm Optimizer for Economic and Environmental Efficiency

In this paper, the ideal two-goal method for systems that must take into consideration environmental, economic, and technological issues is outlined. This strategy aims to improve user satisfaction (UC) by lowering the amount of energy used, the peak-to-average ratio (PAR), the amount of carbon dioxide emissions, and the operational expenses. Through the building of a power distribution system, it is possible to obtain a more efficient usage of energy by households. A beta probability density function (PDF) that is connected to the speed of the wind and solar radiation is utilized in order to display the resources that the sun possesses. The utilization of the proposed optimizer that is based on Deep Reinforcement Learning (DRL), in conjunction with environmentally acceptable energy sources and an effective energy storage system, resulted in a reduction in cost, pollution, user discomfort, and energy consumption. The DRL model is enhanced by the RSOA, which does this by picking the most appropriate values for the target function. Through the use of MATLAB simulations, it was demonstrated that the method was successful. The first scenario clearly demonstrates a large decrease in the pricing of electricity, the PAR, and the carbon emissions when compared to the previously unplanned instance. A demonstration of the effectiveness of the DRL-based strategy is provided here.