良好的飛航排程，除考量航空公司本身的機隊供給及相關作業措施，亦必須兼顧旅客的反應。以往雖有少數文獻於中長期規劃中考量市場競爭之旅客需求變動，然而一般短程飛航排程之研究多以草擬班表為輸入，並假設旅客需求己知且固定，忽略旅客搭機的選擇行為，致使所求得之排程結果未能反映真實之需求狀況，而降低營運績效。緣此，本研究於考量市場需求與該航空公司排程相關性下，建立一變動需求(variable demand)之短期飛航排程模式，以幫助航空公司求得最佳的機隊排程與班次表。本研究利用網路流動技巧構建模式。此模式將包括多重人流與機流網路。在人流網路中，加入旅客偏好選擇模式以定式此一旅客流動問題。為考量等待旅客在實務中的流失情況，本研究使用一般化網路以定式旅客在時空中的變動。在機流網路的設計上，則以整數流動方式定式機隊於時空中的排程。二網路流動間再加上實務限制，以符合實際的飛航作業。本研究模式期能於未來實務應用上，提供為航空業一短期排程暨班次表建立之輔助規劃工具。 本研究以數學規劃方法定式，此模式為一非線性混合整數規劃問題，其屬於NP-hard性質的問題。此問題在求解上更難於以往整數線性的排程規劃問題，因而本研究建立一反覆求解架構，以重複修正該航空公司的市場分配需求，並配合求解固定性需求之短期飛航排程問題，以求解模式。本研究以C電腦語言撰寫演算法，並配合CPLEX數學規劃軟體進行求解。最後，本研究以一國籍航空公司之國內客運營運資料進行一實例測試與分析，再依分析之結果提出結論與建議。 The setting of a good flight schedule for an airline not only has to consider its fleet and related supply, but also has to take into account of passenger reactions on its service. Although little research of medium/long-term flight scheduling in the past has ever dealt with variable passenger demands considering market competitions, almost all past short-term flight scheduling models assumed passenger demands as fixed and used a draft timetable as input to produce the final timetable and schedule, neglecting passenger choice behaviors among different airlines in practice. As a result, the schedule and fleet route offered may not reflect the real demands, decreasing the system performance. Considering both fleet supply and market demands, in this research, we developed a short-term flight scheduling model with variable demands, in order to help an airline solve optimal fleet routes and timetables. We employed network flow techniques to construct the model which includes multiple passengers and fleet flow network. In the passenger flow networks, we introduced a passenger choice model to formulate passenger flows. Considering the loss of waiting passengers in practice, we used generalized networks to formulate passenger flows in terms of time and space. In the fleet flow network, we used integer flow networks to formulate the aircraft routes in terms of time and space. Some side constraints were sat between the passenger and fleet flow network according to the real operating requirements. The model is expected to be a useful planning tool for airlines to determine their short-term fleet routes and timetables. We used mathematical programming techniques to formulate the model as a nonlinear mixed integer program that is characterized as a NP-hard problem and is more difficult to solve than traditional flight scheduling problems that are often formulated as integer linear programs. To efficiently solve the model with practical size problems, we developed an iterative solution framework, in which we repeatedly modify the target airline market share in each iteration and solve a fixed-demand flight scheduling problem with the assistance of the mathematical programming solver, CPLEX. To evaluate the model and the solution framework, we performed a case study using real operating data of domestic passenger transportation from a major Taiwan airline.
