隨著消費習慣的改變，物流運送方式也在不斷變化。網路銷量逐年增加，帶動了運輸需求的提升。傳統的B2B運送模式轉變為B2C或C2C運送，導致傳統路面運輸難以滿足未來的需求。因此，為了提高運送效率和降低成本，未來物流業將更加依賴多元化的運輸工具，其中無人機成為主要的運送方式之一。面對日益增長的物流需求，無人機物流業者需要擴大無人機隊伍並增加服務項目。利用靈活的集合式無人機來進行運送，不僅可以提升服務品質，增加調度靈活性，避免單一服務的問題。目前，國內已有多家業者嘗試使用無人機進行物流運送。如果能事先針對集合式無人機隊進行規劃，不僅能應對需求增長帶來的調度挑戰，還能提升整體服務水準。 本研究構建了一種集合式無人機載貨運送模式，以無人機營運業者的角度出發，使用顧客預約方式，在考慮預約需求與相關無人機運送限制的前提下進行研究，利用數學規劃和時空網路流動技術，以最小化營運成本為目標。該模式可定義為一個含額外限制的整數網路流動問題，屬於NP-hard問題。然而，當處理大型實務問題時，往往因規模過大難以在有限時間求解。因此，本研究結合拉氏鬆弛法與CPLEX，發展了一種啟發式演算法以有效解決問題。為了評估該模型的實用性，進行了範例測試並進行了關鍵參數的敏感度分析。研究結果顯示，該模型及其啟發式解法表現良好，能夠有效降低營運成本並滿足預約需求，對無人機業者未來的運送排程規劃具有參考價值。 ;As consumer habits evolve, so do the methods employed to deliver goods and services. The growth of online sales is driving up transport demand on an annual basis. The traditional B2B delivery model has shifted towards B2C or C2C delivery. This has created challenges for traditional transportation in meeting future demands. Consequently, to enhance the efficiency of delivery and reduce costs, the logistics industry of the future will increasingly rely on a diversified transportation system, with drones becoming one of the principal delivery methods. In light of the mounting demands placed upon the logistics industry, drone logistics operators must expand their drone fleets and enhance their service offerings. The utilization of flexible collective drones for deliveries can not only improve the quality of service and facilitate scheduling flexibility, but also circumvent the issue of single-service dependency. Several domestic operators are currently engaged in experimental operations in drone logistics. If collective drone planning is conducted in advance, it can address the scheduling challenges posed by increasing demand and improve overall service levels. This study constructs an integrated drone cargo delivery model from the perspective of drone operators, focusing on customer reservations and considering both reservation demands and relevant drone delivery constraints. The research employs mathematical programming and space-time network flow techniques to minimize operational costs. The model is defined as an integer network flow problem with additional constraints, classified as NP-hard. However, when dealing with large-scale practical problems, it often becomes challenging to solve them within a limited time due to the scale. Consequently, this study combines Lagrangian relaxation with CPLEX and develops a heuristic algorithm to address the problem effectively. To evaluate the model′s practicality, example tests and sensitivity analysis of key parameters were conducted. The results indicate that the model and its heuristic solution perform well, significantly reducing operational costs while meeting reservation demands, and providing valuable insights for future drone scheduling planning for operators.
