Optimized Energy Management in Renewable Energy Integrated Microgrids Using Pyramidal Dilation Attention Convolutional Neural Networks

The use of Diesel Generators (DGs) as a form of backup puts a stress on the operational cost because of the fuel costs and the levels of maintenance. Furthermore, during periods when demand is high, or renewable supplies are low, system efficiency could reduce overall, meaning greater reliance on less efficient supplies and ultimately greater total loss, particularly where battery storage may not be effectively sized to match demand. To overcome these drawbacks, this manuscript proposes an approach for optimizing the Energy Management System (EMS) in Microgrid (MG) integrated with Renewable Energy Sources (RESs). The proposed method is Pyramidal Dilation Attention Convolutional Neural Network (PDACNN). The primary objective of the proposed method is to minimize the total operational costs and improving efficiency of the system. The proposed PDACNN predicts energy generation and consumption patterns, enhancing the efficiency of the MG system incorporating RESs. Then, the proposed method is implemented in MATLAB and compared with various existing approaches like Flying Foxes Optimization-Deep Attention Dilated Residual Convolutional Neural Network (FFO-DADRCNN), Whale Optimization Algorithm and Pattern Search (WOA-PS), Improved Gradient-Based Optimization-Particle Swarm Optimization (IGBO-PSO), African Vultures Optimization Algorithm (AVOA), and Bald Eagle Search Optimizer (BESO). The proposed PDACNN method achieves an operational cost of $1,837 and an impressive efficiency of 98.7%, demonstrating that this approach significantly outperforms existing methods in both cost-effectiveness and operational efficiency, making it an optimal solution for EM in MGs integrated with RESs.