| 摘要: | 當道路新建完成，維持其服務水準於最佳狀態，使民眾通行無阻兼安全無虞，為道路養護的重要宗旨。除了日常每日例行巡檢養護的災害預防，當遭遇天然災害或其他突發事件時，第一時間需即時應變進行緊急修補的道路搶修小組更是從中扮演重要的角色。他們需要在最短的時間內完成臨時修補，以防坑洞隨著時間與使用持續擴大，造成額外的損害，而後續才會再經由其他養護工程內容進行完整的道路鋪面。過去雖然已經有研究針對緊急搶修工作隊進行探討，但多半著重於道路阻斷限制與後端物資配送最佳化之探討，與道路路面搶修小組同時扮演著搶修工作隊與物料需求補給之角色特性不盡相同，故其模式難以運用在道路路面搶修排程問題。
緣此，本研究利用時空網路流動與數學規劃，以系統最佳化的觀點，在總營運成本最小化目標下，考量實務上相關營運限制，以平均旅行時間構建一確定性道路路面緊急搶修模式。另外，本研究可定義為一含額外限制之整數網路流動問題，屬於NP-Hard問題，為了有效求解實務上大型問題，發展一啟發解法以求解問題，藉由分解問題之觀念，配合CPLEX數學規劃軟體進行模式求解。最後，為了測試本研究模式之實用績效，本研究以第一養護工程處中壢工務段為例，並參考2017年發生的尼莎與海棠雙颱風實際案例，設計三項不同規模大小之範例進行測試與分析，進而提出結論與建議。;When built a new road, maintaining its service level and make road users pass unimpeded is the most important purpose of road maintenance. In addition to the disaster prevention of daily routine inspection and maintenance, when encountering natural disasters or other emergencies, the work team that deal with emergency is also plays an important role. They need to complete temporary repairs in the shortest time to prevent the potholes from continuing to expand over time and use. Most of the emergency repair work schedule in the past were focus on the road blocking constraints and the optimization of logistical support, however, those properties are not similar with this case. In this case, the road emergency repair work team is playing the role of emergency repair team and logistical support at the same time, therefore, the model is difficult to apply to this case.
Hence, we employ network flow techniques, with a systemic optimization perspective and the objective of minimizing the total system cost, as well as the related operating constraints, to construct a deterministic road emergency repair scheduling model. Furthermore, this model is formulated as an integer multiple-commodity network flow problem which is characterized as NP-hard. In order to efficiently solve realistically large problems occurring
in practice, we use a problem decomposition technique a, coupled with the use of a mathematical programming solver CPLEX, to develop a heuristic algorithm. Then, the numerical tests are performed using the first maintenance office data associated with Taoyuan City , Taiwan. Referring to the actual case of the double typhoons Nisha and Begonia in 2017, three case study of different sizes were designed for evaluate the model and the solution algorithm in practice. At last, conclusions and suggestions will be given. |
