當天然災害發生時，如何以最短時間儘快撤離該地區民眾為政府首要任 務。目前疏散車輛的調度與路線規劃，由防災中心的指揮官決定，此種方式 不僅沒有效率，亦可能因決策者的誤判，導致延誤救災情形發生。緣此，本 研究透過時空網路流動技巧，定式車輛及民幕在時空中的疏散情況，並構建 一防災避難疏散排程規劃模式，提供決策者作為在最短時間内疏散並撤離民 眾之有效規劃工具。由於本研究模式為含額外限制之整數網路流動問題，屬 於NP-hard問題。為有效率地求解實務的大規模問題，本研究發展一啟發解 演算法。最後，本研究以國内某一城市可能發生災害之高風險地區的疏散作 業為測試範例，其結果顯示，本研究模式與演算法可為實務應用之參考。

When a natural disaster occurs, the most important action for the government is to evacuate the victims as quickly as possible. In practice, decision makers arrange evacuation vehicles and decide on the escape routes based on their own experience, which is neither efficient nor effective. Thereforet in this research, we utilize a time-space network-flow technique to construct an optimal routing and scheduling model for disaster evacuation. The model is formulated as an integer network flow problem with side constraints, which is characterized as NP-hard. A heuristic algorithm is developed to efficiently solve the large problems that occur in practice. Finally, we perform a case study using real data from a city in Taiwan. The results show that both the model and solution algorithm can be useful references for disaster evacuation in practice. When a natural disaster occurs，the most important action for the government in the duration of damage is to evacuate the victims rapidly. In practice, the decision makers arrange the evacuative vehicles and decide the escape routes based on their own experiences, which is neither efficient nor effective. Therefore, in this research, we utilize a time-space network-flow technique to construct an optimal routing and scheduling model for disaster evacuation. The model is formulated as an integer network flow problem with side constraints, which is characterized as NP-hard. To efficiently solve large problems that occur in practice, a heuristic algorithm is developed. Finally, we perform a case study using real data from a Taiwan city. The results show that the model and the solution algorithm could be useful references for disaster evacuation in practices.