Integrated Energy Management and Load Forecasting Using Machine Learning

Energy management goes beyond just monitoring consumption in modern infrastructure. It involves considering the entire process and creating a balance between demand, supply, and ecological elements. This requires integrated energy management, which is a strategic need for organizations. Integrated energy management enables organizations to combine diverse energy sources, storage systems, and demand-side control to maximize resource usage, reduce costs, and enhance reliability. The main concern for integrated energy management is sustainable energy and the environment. This research emphasizes the role of integrated energy management in maintaining a balance between energy production and consumption, leading to economic sustainability and environmental equilibrium. Machine Learning (ML) algorithms integrated into energy management are the key to significantly improving load forecasting. This study analyzes four well-known models: Random Forest Regression, Multi-Level Perceptron Regression, K-Fold Cross Validation, and Decision Tree Regression. These models were evaluated using real-world data and differ in the level of precision they provide. The Random Forest Regression model is highly accurate, with an R2 score of 0.872, which indicates its strong predictive capabilities. On the other hand, the Multi-Level Perceptron Regression has an R2 score of 0.812, while both the K-Fold cross-validation and the Decision Tree Regression have R2 scores of 0.721 and 0.679, respectively. Based on this information, energy managers can develop their forecasting strategies for electrical network loads.