中文摘要 隨全球人口上昇與有限資源限制，大眾交通工具規劃已成為當前非常值得關注之議題，而在不同區域與條件下，所適合的交通工具發展模式也會有所不同。近年來，自行車租賃系統不斷的在許多城市所發展，因能提供顧客甲地租乙地還的服務，提升了許多便利性。而此類系統在過去經驗當中所會遇到的最大問題，莫過於當顧客到達服務站時，無自行車可提供租賃，或是沒有空檔板可提供還車服務，間接影響的會是顧客滿意度及整個自行車租賃系統的收益，且在每天的交通忙碌時段更為明顯。 在不影響自行車租賃系統原有的模式下，若在需要時使用搬運卡車，依照不同的自行車存放量設定，將自行車重新分配至各個服務站，將會是解決此問題根本的辦法。 本研究以英國Barclays Cycle Hire自行車租賃系統為背景，針對重新分配自行車時之搬運作業，將作業分為Pick-up List Rule（PR）裝載準則、Delivery List Rule（DR）卸載準則和Truck Selection Rule（TR）車輛選取準則三大部分進行探討，並使用Arena以隨機模擬之方式進行模型之建立，而最終將以平均等候時間（Average Waiting Time, AWT）、等候線長度（Average Queueing Length, AQL）及平均服務水準（Average Service Level, ASL）及每日平均行走距離（Average Daily Distance, ADD）進行不同搬運組合之績效評估，並分析在執行搬運作業時影響結果之關鍵因素，與不同績效評估指標間之關係與涵義。
關鍵字：自行車租賃系統、車輛調度問題、多輛搬運車作業、隨機模擬 ;Abstract An increasing global population and limited supplies of natural resources have made public transportation systems a popular choice worldwide. Consequently, the planning for these systems has received increasing attentions recently. The different conditions and limitations in different cities determine whether a transportation model is appropriate or not. In recent years, Bike-sharing System (BSS) has been adopted in many cities all over the world due to the convenience and environmental friendli-ness provided by these systems. However, they often suffer from unbalanced num-bers of bicycles within the system, causing customers difficult to rent or return bicy-cles. Assuming that the number of bicycles does not change in the system, the best approach to solving this problem is to use trucks to redistribute bicycles to each rental station using effective vehicle dispatch rules within a shortest possible time window. We propose an effective vehicle dispatch strategy to reposition bicycles using multiple trucks in three rules (Pick-up List Rule, Delivery List Rule and Truck Selec-tion Rule) and use simulations to create dynamic model. The four performance in-dicators (Average Waiting Time, Average Queueing Length, Average Service Level and Average Daily Distance) will use to evaluate the result after operation and nu-merical analysis will figure out the key factory and relationship between each other.
