Optimized guidance for building fires considering occupants’ route choices

Occupants in building fires usually have difficulties to evacuate safely, and providing guidance may help them select proper routes. However, groups of occupants may not follow guidance because they make their own decisions and are affected by psychological factors. Their reactions to guidance should therefore be considered by an emergency guidance system. Most existing studies assumed that occupants would follow guidance, and congestions might occur if suggested routes were not used. To consider occupants’ route choices, the random utility theory is employed to characterize their reactions to guidance, and the guidance is optimized in a leader–follower​ (Stackelberg) game framework in this paper. To partially validate the effect of guidance and determine related parameters of psychological factors, data from virtual reality (VR) experiments are analyzed, where human participants were recruited and their probabilistic route choices were analyzed. Using parameters derived from VR experiments, numerical and simulation results demonstrate that optimized guidance is effective in terms of the overall risk and evacuation times because related congestions are alleviated. A sensitivity analysis is also performed to assess the impacts of different psychological factors on optimized guidance and evacuations.