我們觀察出當行人在建築物內移動至出口時,大部分行人皆會選擇走離自己最近的出口和可以讓自己用比較輕鬆省力的方式移動到的出口,例如:手扶梯,而如果同時間大量行人陸續往同一個出口移動會出現排隊和擁塞的行為,而且有部分行人會因為觀察出這些形成擁塞的地方,而避開這些擁塞處選擇走另一個較遠或需要花比較多力氣移動的出口,而本研究希望可以藉由建立細胞自動機(cellular automata ,CA)行人模擬重現真實案例行人避開擁塞行為的模式,並且最佳化這套細胞自動機行人模擬模式,讓電腦模擬出來的巨觀行人移動狀況和產生的行為趨勢接近真實案例中行人出現的移動狀況和產生的行為,更進一步我們可以利用本研究所建立的電腦模式預測各類型不同案例中的行人避開擁塞的行為及移動狀況。 We observe that pedestrians tent to leave a building via, the nearest exit in the way that requires less effort. For example, one way for people to save time and energy is using escalators. If many pedestrians move to the same exit simultaneously, there would be congestion. Some pedestrians will observe and avoid the crowded places, and instead go through one of the more distant exits. In this study, we simulate the case where the pedestrians avoid crowded places using cellular automata. We also calibrate the model to reproduce the macroscopic pedestrian movement behaviors. Furthermore, we validate the model with two additional cases.