Research on Machine Learning and AI for Prediction of Energy Consumption, Classification, and Analysis of Energy Conservation Potentials in Buildings

Research on ensemble Forecast modeling for energy consumption and dynamic benchmarking in different sectors of buildings. Through extensive data collection of over 100 buildings, literature review, statistical analyses, and collaboration with various stakeholders, a robust and accurate Forecasting Model was developed through this research. The resulting Forecasting model and benchmarking holds immense potential to drive positive change in the building sector, facilitating energy efficiency, sustainability, and energy conservation.
Benchmarking using machine learning techniques helps classifying the building as Excellent, Good, Average and Bad with respect to specific energy consumption and it is modeled with multiple dynamic factors. There has been numerous research on Benchmarking using regression models and machine learning. But most of the time it is noted that the same is limited to Energy consumption per square meter area. However, this research paper emphasizes a different approach considering numerous variables like Weather and Occupancy. Most of the time these variables are overlooked and for the purpose of ease only energy density is considered for benchmarking parameters like kWh/Sqm as industrial practices