考慮指定席與自由席旅客流動關係之鐵路座位配置模式

A Railway Seat Allocation Model Considering the Passenger Flow between Reserved and Unreserved Seat

高鐵為公共服務業且為追求最大營收，可能使自由席車廂過度擁擠，故須將公平性及旅客轉移量皆納入考量，以限制賣票數量。本研究利用營收管理概念，建構高鐵旅客在指定席與自由席間流動關係之數學規劃模型，並用非凌駕排序基因演算法II求出最佳配位模式，以提供鐵路運輸業配置座位，創造潛在營收。結果顯示，考慮相依需求、旅客流動及公平性且不考慮多費率、取消訂票或未出現及自由席車廂擁擠程度下，將比不考慮公平性之期望總營收提升19.32%，且自由席車廂之擁擠程度無顯著差異，但某些班次有過於擁擠之現象。而若要滿足最大營收及公平性之目標，配置座位時，直達車、每站皆停與半直達車、從台中出發每站皆停列車應分別優先配給長途、短途及中途旅客。 The Taiwan High Speed railway is a public service industry. For pursuing maximum revenue, it mays result in overcrowding of unreserved seats. Therefore, fairness and the amount of passenger transfers must be taken into account in order to limit the number of tickets sold. This study uses the concept of revenue management to construct a mathematical planning model for Taiwan High Speed railway passengers in reserved seats and unreserved seats, and uses Non-dominated Sorting Genetic Algorithms II to find the best model to allocate seats for the railway transportation industry and create potential revenue. The results show that considering interdependent demand, passenger flow and fairness and not considering multiple rates, canceling bookings or no-show and congestion of unreserved seats, will increase 19.32% over the expected total revenue without considering fairness. And there is no significant difference in the degree of congestion in unreserved seats passenger cars, but some trains are overcrowded. In order to meet the goal of maximum revenue and fairness, when seats are allocated, express trains, local trains and limited stop trains, and local from Taichung trains should be assigned to long-distance, short-haul and the middle journey passengers respectively.

The Taiwan High Speed railway is a public service industry. For pursuing maximum revenue, it mays result in overcrowding of unreserved seats. Therefore, fairness and the amount of passenger transfers must be taken into account in order to limit the number of tickets sold. This study uses the concept of revenue management to construct a mathematical planning model for Taiwan High Speed railway passengers in reserved seats and unreserved seats, and uses Non-dominated Sorting Genetic Algorithms II to find the best model to allocate seats for the railway transportation industry and create potential revenue. The results show that considering interdependent demand, passenger flow and fairness and not considering multiple rates, canceling bookings or no-show and congestion of unreserved seats, will increase 19.32% over the expected total revenue without considering fairness. And there is no significant difference in the degree of congestion in unreserved seats passenger cars, but some trains are overcrowded. In order to meet the goal of maximum revenue and fairness, when seats are allocated, express trains, local trains and limited stop trains, and local from Taichung trains should be assigned to long-distance, short-haul and the middle journey passengers respectively.