A coupled evacuation model considering physiological tolerance and dynamic speed adaptation for fire smoke

Smoke is one of the most critical factors affecting evacuation performance during building fires. However, existing models often fail to fully consider its impact on occupant physiology and movement. To this end, this study proposes a coupled evacuation model that incorporates physiological tolerance and dynamic speed adaptation under fire smoke conditions. Physiological tolerance is quantified using the integrated hazard dose (IHD) model. Additionally, the dynamic speed adaptation mechanism adjusts evacuation speed according to environmental conditions. Validation against real fire cases demonstrates that the proposed coupled model can estimate evacuation results more consistently than uncoupled or partially coupled models. Furthermore, XGBoost was further integrated with NSGA-II to support evacuation strategy optimization. The results indicate that evacuation risk can be significantly reduced when the initial occupant load is controlled within an appropriate range. This study provides a practical tool for performance-based fire safety design and offers useful insights into reducing evacuation risk under smoke conditions.