考量交通級聯反應下隧道補強策略之研究：基於路網韌性之觀點

Abstract

臺灣因地理環境特殊，經常受到颱風、地震等天災影響。災害所造成的破壞，往往引發嚴重的交通問題，不僅影響民眾的日常生活與生計，更可能導致救援行動的延誤，進一步加重災情的影響。然而在交通基礎建設中，又以隧道破壞所造成的影響最為嚴重。由於隧道的封閉性，發生災害後不僅僅會導致路網的破碎，由於與一般道路與橋樑不同，無法透過空中進行救援，更是會使的災情更為嚴重。而山多平原少的東部，是臺灣隧道密集區，在災後更是可能出現「孤島」，使的部分地區的民眾無法維持日常生活。因此，本研究探討災後隧道破壞下的級聯反應對於路網韌性的影響。級聯反應是指當一條隧道被完全破壞後，導致流量選擇其他替代道路。而這個行為會導致其他道路上的流量增加，導致總旅行時間的上升。而在過往文獻在評估韌性時，常將路網元件視為是獨立個體，然而路徑的容量會因為瓶頸隧道的容量而有所不同。 本研究主要聚焦於災前的隧道最佳化補強策略問題，並考慮級聯反應帶來的影響並以路網的角度去評估韌性。提出一個雙層的模型，上層是補強策略的最佳化問題，用來決定哪些隧道需要進行補強。而下層是系統最佳化的交通指派問題，用來反映災後的交通流分配，並透過流量的權重來表示級聯反應對整體路網的影響。 本研究使用臺灣國道、省道路網作為案例，並針對2024年4月3日的花蓮地震與2025年1月21日的嘉義地震作為災害情境去進行分析。更進一步去比較有無考慮路徑串聯關係與有無考慮流量之下補強依據之間的差異。最後提出對於預算和災害嚴重程度的敏感度分析。結果顯示，本研究提出的方法 (同時考慮隧道的串聯關係與級聯反應) 可以有效提升路網整體韌性並減少總旅行時間。

Taiwan is located on the Circum-Pacific Belt, where earthquakes frequently occur and pose significant threats to the structure and stability of transportation infrastructure, including roads, bridges, and tunnels. Among the infrastructure, tunnels are particularly critical due to their nature of enclosed environments, which makes rescue and evacuation efforts much more challenging when damage occurs. This often leads to prolonged disaster relief operations and greater cascading effects over the entire transportation system, highlighting the importance of enhancing tunnel resilience against natural disasters. Hence, this research aims to assess the resilience of a roadway network from a holistic perspective and enhance the network’s resilience by reinforcing the most critical tunnels as part of pre-disaster planning. To achieve this, a bi-level optimization model is proposed, considering the reinforcement problem using capacity as an indicator to evaluate resilience. The upper-level model determines which tunnels to be reinforced by maximizing the network resilience. The lower-level model deals with the traffic assignment problem. We adopt the system optimal condition to present the traffic control in the post-disaster. A case study is made over the freeway and provincial roadway networks, and we focus on the tunnels over the eastern corridor of Taiwan. Two earthquake scenarios happened on April 3, 2024, and January 21, 2025 are employed, and comparisons are made over the modeling settings of whether or not the path flow and the series relationship between tunnels are taken into account in the analysis. The results indicate that the proposed method, which simultaneously considers the series relationship and path flow, can effectively enhance the network resilience and reduce the total travel time.