User-Centric Comfort Measurement and Energy Optimization: a Pareto-Efficient Approach for a Personalized Recommendation Strategy

This study introduces a user-centered recommendation system that integrates environmental measurements (both historical and forecasted through artificial intelligence) with subjective thermal sensation data to compute the optimal set-point temperature and humidity levels. The measurement system is based on a Pareto-efficient optimization algorithm, which leverages historical building energy consumption and predicted outdoor temperatures to identify solutions that simultaneously guarantee user comfort and minimize energy consumption. The set-points derived from the algorithm are the core of the recommendation system that suggests environmental settings tailored on occupants' subjective thermal perception. Unlike conventional approaches that often disregard individual thermal perception, this methodology adapts to user-specific thermal preferences, ensuring that the recommended conditions align with personal comfort while improving energy efficiency. To validate the methodology, a case study was conducted in an office building in Finland, equipped with a comprehensive set of IoT sensors. Results show that applying the recommended conditions leads to a 44.8% reduction in energy consumption while ensuring comfort to occupants for the overall stay in the building. These findings underscore the importance of user-centered comfort measures in energy-efficient building management.