Experimental Assessment of Indoor Air Quality and Thermal Comfort Performance Using Fuzzy Logic Control for Natural and Mechanical Ventilation Systems

Ventilation strategies embody a fundamental compromise between human comfort, indoor air quality, and the responsiveness of building systems to dynamic occupancy. As modern buildings become increasingly airtight, the ability of ventilation systems to adapt in real time has become essential. This study presents an experimental comparison of natural ventilation (NV) and mechanical ventilation (MV) operated under an identical two-level hierarchical fuzzy logic control, coupling thermal comfort assessment (temperature and relative humidity) with CO2-based ventilation decision-making. Experiments were conducted in a four-room testbed under equivalent occupancy profiles, aligned over 24 hours to ensure comparability. Indoor air temperature, relative humidity, and CO2 concentration were monitored to evaluate thermal stability, indoor air quality performance, and ventilation effectiveness. Both strategies maintained indoor temperatures close to setpoints (tracking errors < 2.5 °C), but NV produced larger short-term fluctuations in temperature and humidity. MV achieved better indoor air quality control, limiting CO2 peaks to typically below 1200 ppm, whereas NV reached higher peaks (up to ~1400 ppm in higher-occupancy rooms). Post-occupancy CO2 decay analysis revealed faster and more consistent pollutant removal under mechanical ventilation, compared to lower and more variable decay rates under natural ventilation. Further exposure analysis showed that mechanical ventilation increased the proportion of time spent below 1000 ppm by up to 60% across rooms. Overall, the results indicate that applying a common fuzzy decision logic allows both ventilation strategies to operate effectively, while mechanical ventilation provides more stable and predictable indoor environmental control. Natural ventilation, although viable, exhibits greater variability and stronger room-dependent behavior, highlighting the potential of hybrid strategies that combine both approaches.