隨著都市化與經濟穩定發展下，都市交通網路連結趨於密集，國際上主要城市皆致力於解決交通問題與改善市民生活品質，近年來仰賴RFID普及使市民們能快速進行電子交易，同時全球環保意識抬頭使市民們嚮往低碳、樂活生活圈，以上皆是促使公共自行車租賃系統成為國際上熱門公共計劃。公共自行車與以往租賃系統不同，在於使用者入會簡便且能快速借、還自行車，除此之外各個自行車站皆不需要安排人力進行租借服務，多數城市都實施短程租借免費或優惠，公共自行車租賃系統普遍被視為大眾交通一環並連結城市交通樞紐，公共自行車其公共性質遠多於商業性質，設置地點也不同於以往設置在觀光景點而是完全融入城市生活圈，公共自行車系統將做為市民通勤、休閒、便利與大眾交通的最後一哩路。 近年來公共自行車系統已成為都市民眾生活的一部分，不僅增加城市形象與更成為疏通交通的工具，雖然公共自行車站不需要人力協助租借服務，但仍會出現定期維修與車輛調度之需求之服務需求，所以仍需要人力維持服務品質。當城市仰賴公共自行車租賃系統，則需要持續維持高服務水準與服務品質，此時如何有效提升服務率變成十分重要課題。公共自行車租賃服務與商品出售不同，因為交通網路始終維持固定數量自行車，搬運車補貨必定來自其他自行車站，且過多與過少配置都會形成浪費，因為需求本身無法準確的預測，故有必要探討如何有效有限運用資源產生最大經濟效益。 本研究探討車輛調度因子組合以此提升服務效率，以英國Barclays Cycle Hire實際情境設計實驗，針對選取搬運位置與卸貨位置之組合促使搬運車對自行車站進行庫存調節，利用統計分析探討在四種績效指標下之最佳車輛調度因子組合歸納出MEP搭配NPD等候時間、長度以及搬運車行走距離都有較佳表現，另外以MQSP、MAWTP、MFIFOP裝載因子與MSD、MSLD卸貨因子在成功借車率上有較好之表現。 ;With the steady urbanization and economic development, the urban transport network links tend intensive. In recent years public-bike sharing system has become internationally popular public programs. Public bike rental system with the previous difference is that membership can quickly rent and easy become membership. Similarly, Public bike rental system location is also different from the previous system in the tourist attractions, but integrated into the urban. Most important thing, public-bike sharing system will be used as people commuting, leisure, convenience and transportation last a mile. Public bike rental service station does not need human assistance, but there will still needs regular maintenance schedule of the vehicle. Public bike rental system is different from sale of goods, because the network has remained fixed number of bicycles and replenishment must come from other bike stations. Therefore, it is necessary to explore how to effectively generate maximum economic benefits in limited resources. In order to enhance vehicle scheduling efficiency this study investigated the combination of factors. In the experimental design use the actual situation and statistical analysis to explore the optimal vehicle dispatching under four performance indicators. Combination MEP rule and NPD rule perform better than other factor level under queueing length, waiting time, daily distance performance. Additionally, MQSP, MAWTP, MFIFOP rule with MSD, MSLD rule perform well in service level performance.
