後疫情時代，無人機在物流運送領域的應用正迅速發展，國外已出現許多商業上的應用案例，各國亦投入許多資源致力於無人機的技術發展，顯示無人機於物流運送領域的發展是強力且不可逆的，在可預期的未來，無人機將成為物流運送的主要運具之一。因此，為了因應大量的物流運送須求，未來的無人機物流業者勢必得擴大無人機隊規模及增加服務項目，且依不同任務需求，業者可以使用不同的無人機進行運送，除了達到更完善的服務外，亦可透過更專業的分工來降低整體營運成本，增加調度的靈活性，避免落入服務過於單一的窘境，而目前國內已有許多業者正嘗試利用無人機進行物流運送，因此若能預先針對一混合不同機型的無人機隊進行物流運送排程規劃，不僅可以應對需求量增加後無人機的運作調度，擴大服務項目後亦可提升服務水準，帶來業者與使用者的雙贏。 本研究構建一混合無人機隊物流運送模式，以無人機物流業者的角度，在考量電量限制及實際運作飛行時的相關限制，並以所有任務均須完成之前提下，針對一日之營運進行無人機的運送排程規劃。模式應用時空網路流動的觀念，以每日營運最小成本為目標，透過數學規相關理論依據，建構一混合無人機隊物流運送模式，模式中除包含流量守恆限制外，亦加上一些額外的限制如電量限制等，以滿足實務的營運條件。由於問題規模龐大且屬NP-hard問題，因此本研究透過拉氏鬆弛法配合CPLEX，發展一啟發式演算法。並為了評估模式與演算法之可行性與績效，以隨機方式產生不同規模之測試範例，進行測試與分析，最後針對重要參數進行敏感度分析進而提出結論與建議。 ;Drones have witnessed significant advancements in the field of logistics following the COVID-19 pandemic, demonstrating their increasing proficiency in this domain. The widespread adoption of drones for business purposes by numerous countries highlights their pivotal and unstoppable role in logistics. It is evident that drones are poised to become indispensable for transportation in the near future. To effectively meet the growing demand for logistics services, drone logistics operators must expand their fleets and diversify their service offerings. By leveraging different types of drones for transportation, operators can not only provide a wider range of services but also optimize task allocation, save costs, streamline scheduling, and mitigate over-specialization. Currently, many operators have already commenced the utilization of drones for logistics transportation. Planning and scheduling a diverse mix of drones with varying models in advance can not only ensure seamless coordination among drone operations to meet the surging demand but also enhance the overall service quality, benefiting both operators and users.
This study adopts the perspective of drone logistics operators and aims to develop a comprehensive model for mixed drone fleet logistics transportation. The model takes into consideration crucial factors such as battery limitations and real-world flight conditions. The primary objective is to devise a scheduling and planning model that encompasses all the tasks involved in daily operations. By drawing upon concepts from space-time network flow, the study strives to minimize daily costs using mathematical theories and incorporating practical constraints such as flow conservation and battery limits to accurately emulate real-world conditions. Given the inherent complexity of the problem, which falls under the category of NP-hard problems, the Lagrangian relaxation method and CPLEX are employed as solution strategies. To assess the effectiveness of the proposed model and solution approach, a variety of random test cases of different sizes are generated for rigorous analysis and testing. Additionally, sensitivity analysis is conducted on crucial variables to obtain insightful results and recommendations, thereby refining the model′s performance and suggesting optimal strategies for mixed drone fleet logistics transportation.
