AI-Assisted Energy Consumption Forecasting for Households in Negros Occidental

Accurate energy forecasting in residential areas is key to promoting sustainability, reducing energy costs, and enhancing grid reliability. This study presents an AI-assisted energy consumption forecasting model designed specifically for households in Negros Occidental, Philippines. Leveraging an ensemble model that combines a Deep Neural Network (DNN) and an XGBoost regressor, the research aims to generate precise energy forecasts using variables such as climate, household size, and appliance usage. The model utilizes historical energy data, smart meter inputs, and environmental conditions to train and evaluate performance. By emphasizing both accuracy and explainability, the hybrid model addresses the need for user-friendly energy solutions tailored to socio-environmental contexts in developing regions. Additionally, the study explores the relevance of localized forecasting in the face of frequent brownouts and the growing adoption of solar technologies. The results demonstrate the superiority of the ensemble approach over individual models, validating the method’s applicability in real-world scenarios. This research contributes a scalable, transparent, and adaptable solution that aligns with national sustainability goals and encourages household-level participation in energy management.