- 學位論文
快捷巴士班距營運穩態系統開發
Development of BRT Operation System with Frequency Stability
若您是本文的作者，可授權文章由華藝線上圖書館中協助推廣。
摘要
快捷巴士系統(BRT) 具備公車系統(Bus)以及大眾捷運系統(MRT)之優點，其建設經費較少、施工期程較短之，又具有較大運能，可在短期內改善城市之交通問題，亦可做為未來捷運系統推動之基礎。如今BRT成為世界各城市改善城市交通的重要選擇，因此台中市政府發展快捷巴士以改善大台中地區的交通問題。 本研究目的為透過速率的調整，研擬策略維持快捷巴士的班距穩定，增加快捷巴士的可靠度，以提昇民眾使用快捷巴士的意願，進而幫助快捷巴士系統的營運，並降低快捷巴士路線對台中市路網的負面影響。本研究針對快捷巴士系統的兩個績效評估指標，準點率以及車間距，研擬維持與改善的策略，解決快捷巴士可能會面臨的誤點問題以及群聚現象，並且降低快捷巴士停等紅燈的機率，減少優先號誌的使用頻率。 本研究使用兩個策略，分別為準點追分策略以及穩定車間距策略。準點追分策略包含一般追分模式以及預測駕駛特性追分模式兩模式。穩定車間距策略包含預測駕駛特性模式以及延長綠燈模式，本策略兩模式可以配合使用，故有四個模式組合，分別為不做預測與不延長綠燈(NN)、僅使用延長綠燈(NG)、僅使用預測駕駛特性(AN)、預測駕駛特性配合延長綠燈(AG)。穩態系統策略的使用順序為，當誤點發生時即啟動準點追分策略，而當快捷巴士使用準點追分策略到達下一停靠站時，仍然與時刻表時間產生誤差，即定義為追分失敗，此時將啟動穩定車間距模式，維持前後兩車之車間距。 兩策略所使用的預測演算法是類神經網路，其預測輸入的資料包含誤點時間以及期望旅行速率，而快捷巴士則根據類神經網路預測的速率進行操作。由於目前快捷巴士尚未正式營運，因此構建類神經網路的資料需透過VISSIM模擬取得，而模式的績效驗證亦在模擬環境下運作。 本研究選定快捷巴士藍線A09至A10站間距對兩策略進行績效評估。準點追分策略根據MAPE值得比較，預測駕駛特性追分模式(0.0473)績效優於一般追分模式(0.0961)。而在穩定車間距策略方面，MAPE值由小至大分別為AG(0.0518)、AN(0.0704)、NG(0.1541)、NN(0.197)。停等紅燈機率由低至高依序為AG(2/72)、AN(5/72)、NG(7/72)、NN(15/72)。
並列摘要
BRT possesses the benefits of bus and MRT. It costs less, requires less time to construct, has more capacity to alleviate traffic congestion in a short period, and can be used as a foundation for MRT system in the future. So far BRT has become an important solution for solving traffic problems in cities around the world. Therefore, Taichung City Government developed BRT to improve traffic conditions in Taichung. The goal of this study is to boost citizen’s willingness to take BRT, assist in its operation, and to reduce the negative effects BRT has on the transportation network in Taichung by developing strategies to maintain the stability of frequency of BRT to increase its reliability. The study targets at two performance indexes, punctuality and headway, to plan strategies for maintenance and improvement to deal with the delay and bus bunching that may occur, and to lower the possibility of stopping at the traffic lights and the frequency of using priority signals. The study makes use of two strategies, which are Schedule-Based Control and Headway-Based Control. Schedule-Based Control comprises “Normal Schedule-Based Model” and “Driver Characteristics Prediction Schedule-Based Model”. Headway-Based Control includes “Driver Characteristics Prediction Model” and “Green Light Extension Model”. The two models in this strategy can be used in combination so there are four scenarios, which are no driver characteristics prediction nor green light extension (NN), green light extension only (NG), driver characteristics prediction only (AN), and driver characteristics prediction with green light extension (AG). The sequence of implementing of strategy is: Implement Schedule-Based Control when a delay occurs. If the BRT’s arrival time deviates from the timetable (i.e. an “error” occurs) after implementing Schedule-Based Control, we define this situation as failure of schedule maintain. We will then implement Headway-Based Control to maintain a certain headway between two buses. The predictive algorithm employed in these two strategies is “Artificial Neural Network”, whose data input consists of delay and expected travel speed. BRT conducts manipulation according to the speed calculated by Artificial Neural Network. Since BRT has not yet been officially operated, the data required for building Artificial Neural Network need to be gathered using VISSIM, and the examination of the model’s performance is conducted in a mock environment. We chose A09 and A10 on BRT blue line to do the performance valuation. Schedule-Based Control can be compared according to MAPE figure. Driver Characteristics Prediction Model’s performance (0.0473) is superior than Normal Schedule-Based Model (0.0961). For Headway-Based Control Strategy, MAPE figure, presented from the smallest to the largest, are AG(0.0518), AN(0.0704), NG(0.1541), NN(0.1970). The possibility of stopping at the traffic light, ranked from the lowest to the highest, are AG(2/72), AN(5/72), NG(7/72), NN(15/72).
參考文獻
14. 王信傑(2006)，「輕軌路口設置型式準則之研究」，臺灣大學研究所碩士論文。
23. 鼎漢國際工程顧問有限公司(2011)，「中台灣公車捷運系統(BRT)服務規劃設計與可行性評估」。
24. 劉姿君(2013)，「基於號誌因子之公車動態旅行時間預估模式研究」，臺灣大學研究所碩士論文。
1. Aashto, A. (2001). Policy on Geometric Design of Highways and Streets.American Association of State Highway and Transportation Officials, Washington, DC, 1, 990.
2. Chien, S. I. J., Ding, Y., & Wei, C. (2002). Dynamic bus arrival time prediction with artificial neural networks. Journal of Transportation Engineering, 128(5), 429-438.