Towards Efficient Building Energy Management: Deep Learning and Iot Strategies for Predictive Analytics and Optimization

Abstract

To accomplish the aim in day to day life for the worldwide increased population, Smart Building Energy Management System (SBEMS) assures a significant part by using automated process. In this study, a Deep Learning and IoT based Smart Building Energy Management approach is proposed to predict the energy consumption, categorize the building into demand-wise zone and to maintain and monitor the electrical equipment efficiently and smartly. Previously, various traditional statistical methods have been used on the classical time-series data for load forecasting. In this research, some advanced Machine Learning algorithms are investigated to be performed over the data-set for the prediction and efficient management of load for better accuracy. To cluster the low, medium, and high energy demand zone of a building, K-means clustering algorithm has been proposed in this research paper that ensures the maximum and proper utilization of energy. In the previous research works, limited studies have been conducted by gathering all the important features in one umbrella and focusing on the appropriate approaches of load forecasting, clustering and anomaly detection with the unique data-set. However, our research work shows a detailed Building Energy Management Analysis by overcoming this limitations with better accuracy.

Keywords: Smart building energy management system (SBEMS), Deep learning, IoT, Load forecasting, Clustering, Anomaly detection