本研究針對臺鐵線上訂票現況進行資源重新分配的檢討，透過預購票的預訂時間與數量控制的策略研擬與評估，以改善臺鐵的票務管理與經營績效。研究中假設三種顧客（即旅行業者、商務人士與一般人士）的訂票行為，藉由系統模擬軟體 Arena 模擬九種管理策略（三種取票時間方式：兩、三與五天，並搭配三種釋票策略：全開釋放、兩階段釋放、與每天釋放 10%），以 評估不同的策略組合的績效。在模擬實證分析中，本研究藉由搭乘人數、退票人次、取消票次與訂票失敗數的情境設計，搭配不同目標函數（包括：最大化收益、最小化退票成本與最大化總淨值）來衡量不同票務管理策略的績效。研究結果顯示，以最大化總淨值為目標的情況之下，最佳的策略為取票日五天與分兩次釋票的方式；若以最小化退票成本為目標，則取票時間五天與每天釋放 10%的釋票方式為最佳的售票策略。

This research investigates the problems associated with the Taiwan Railways Administration’s （TRA’s） on-line ticket booking system which does not allow for price discrimination and overbooking. We therefore focus on the issues of redistribution strategies for the limited ticket resources. In this research, we construct a simulation model and assume that there are three kinds of customers （i.e., travel agent, businessman and general customer） each with his or her individual booking behavior in the TRA’s on-line ticket booking system. Nine scenarios combined different booking confirmation deadlines （i.e., two days, three days and five days） and various resource ticket released strategies （i.e., full allowance in the beginning, half allowance in the beginning and full allowance after a week, and constant daily allowance rate - 10%） are analyzed via simulation experiments with the simulation software Arena. System performances are evaluated by three different objectives （maximizing total system revenue, minimizing total ticket cancellation costs, and maximizing total system net value）. Findings in this study show that the best ticket management strategy in terms of maximizing total system net value is to set the booking ticket confirmation deadline as three days after booking combined with the resource ticket released half allowance in the beginning and full allowance after a week. If decision makers aim to pursue the minimal total ticket cancellation costs, setting the booking ticket confirmation deadline as five days with the resource ticket released 10% allowance daily would be more desirable.