Real data-driven occupant-behavior optimization for reduced energy consumption and improved comfort

A significant amount of energy can be saved through improving occupant behavior. However, implementing energy-saving behavioral changes requires careful consideration based on real-life data to avoid sacrificing comfort. Towards addressing this need, this paper proposes a real data-driven method to assess the potential of occupant-behavior improvements in simultaneously reducing energy consumption and enhancing comfort. The proposed method consists of two main components: (1) machine learning-based occupant-behavior-sensitive models for real data-driven prediction of cooling and lighting energy consumption and thermal and visual occupant comfort; and (2) a genetic algorithm-based optimization model, which uses the machine-learning models to compute the energy consumption and occupant comfort and accordingly optimizes occupant behavior for reduced energy consumption and improved comfort. The proposed method was tested on real data collected from an office building. The experimental results showed potential behavioral energy savings in the range of 11-22%, with a significant improvement in occupant comfort.

Automated summary

This study proposes a real data-driven method using machine learning and optimization models to evaluate and optimize occupant behavior for reducing energy consumption and improving comfort.