Application of random forest algorithm in identification of key factors in building cooling system energy consumption and energy conservation strategies (for office-educational buildings in Amirkabir University of Technology)

Poursedigh, Romina; Hatami, Farzad; Shakeri, Eghbal

doi:10.22060/aest.2025.5751

Application of random forest algorithm in identification of key factors in building cooling system energy consumption and energy conservation strategies (for office-educational buildings in Amirkabir University of Technology)

Document Type : Original Article

Authors

Romina Poursedigh ¹

Farzad Hatami ²

Eghbal Shakeri ¹

¹ Civil and Environmental Engineering Faculty, Amirkabir University of Technology

² Structural and Earthquake Research Institute, Amirkabir University of Technology

10.22060/aest.2025.5751

Abstract

The increasing electricity consumption for air cooling and conditioning in buildings has become a significant challenge in Iran, driven by population growth, global warming, limited energy production and distribution capacities, and concerns about the reliance on fossil fuels for electricity generation. This study investigates energy consumption patterns of cooling systems in three buildings at Amirkabir University of Technology: Civil and Environmental Engineering (Building No. 2), Computer Engineering, and Aerospace Engineering. Using energy consumption data, the study identifies key factors related to building characteristics and user behavior that impact cooling energy usage. A Random Forest analysis revealed that among 14 factors, 9 were most significant, with the number of staff, number of students, and the distribution percentage of operating units ranked as the top three. These findings provide insights for developing effective energy conservation strategies tailored to university buildings in similar contexts

Keywords

Building Energy Consumption

Optimization

Random Forest Algorithm

Occupants’ Behavior

Sustainable development