本文主要探討以優先號誌控制方式縮短行車時間，藉以提升公車準點到站之機率。研究中選定台南市5路公車於西門路上連續三處號誌化路口所組成之四區段作為研究區域。根據該路段範圍內之道路幾何、起訖旅次及交通管理與控制等資料，建立PARAMICS車流模擬軟體之相容格式檔，此路網之車流模擬結果則用以建立其公車基礎班表及各區段之行車資料，以作為公車準點性及行車時間預估依據。文中採用PI控制理論來預測公車下一時間速率及到達下一個路口所需之時間。預測結果顯示，此方法可於公車到達路口前5秒即可準確預估其到達時間之特性，因此將有足夠之時間判斷路口執行號誌優先控制之必要性及所採取之模式。根據PARAMICS車流軟體之模擬資料，本研究建立PI控制法用於行車時間估算方法，經1200次估算之結果與PARAMICS數據具有相似之變化。文中利用該行車時間估算方法來發展路口號誌之優先控制策略，其方式係根據行駛期間之預估到達路口時間資訊，來判斷公車是否準點到站，並作為執行號誌控制之依據。
研究中係採用延長綠燈時間及縮短紅燈時間之號誌控制方法；為達路口效益最大化及公車優先通過之目標，本文係採綠燈延長或紅燈縮短時間不大於6秒之控制策略參數，而控制目標式則為兩站間的旅行時間，且不大於68秒，此值當在沒有號誌優先通行控制時，僅有30％的公車可準時或提前到站。文中之三處路口號誌共以八種控制模式進行設計，依序由綠燈延長、紅燈縮短時間為3秒來執行各控制模式之模擬；若行車時間已達其設定值，即執行下一發車時刻之模擬。否則即增加控制時間至6秒。採每秒發車共6300次模擬，當三處路口之控制時間均為6秒時，因執行綠燈延長及紅燈縮短控制，兩站間之公車行車時間低於68秒之車次，將分別增加為40％到50％，故此可驗證號誌優先控制將提升公車到站之準點性。
雖然本研究中僅分析三處號誌化路口之案例，但相同之控制邏輯仍可應用於較長路段之多號誌路口之控制場合；另外由於GPS之車行資料可隨時更新車行時間預估軟體之資料庫，因此隨著號誌優先通行控制次數之增加，預估公車到站時間之可靠度及路口號誌控制之有效性均可有效地提高。

This study aims at developing new signal preemption control strategies in order to enhance the timetable adherence of city bus lines within bus stop pairs. Three adjacent signalized intersections in No. 5 bus line in Tainan City are selected for doing a case study. All data of the its roadway network, O-D trip matrix and traffic control measures within the said study area are put into PARAMICS simulation software for further analysis. Simulation results are employed to establish the basic bus timetable which is based for evaluating the prediction accuracy of bus travel time as well as bus timetable adherence. Using PI control with 1,200 simulation runs, the proposed travel time prediction model demonstrates quite satisfactory performance compared with those simulation outputs.
The travel time prediction model is then adopted to develop signal preemption control strategies and is integrated with each other for following analysis. Two primary signal priority schemes, i.e. the extension of major green period and the reduction of major red period are employed to enhance the possibility of bus priority pass through those intersections for maximizing the benefit of bus timetable adherence. Eight control modes comprised by three adjacent signalized intersections have been designed to engage in simulation tests. A bus releases in every second and is traced throughout the bus network until it leaves out the system. Total simulation time lasts 6,300 seconds. It is found that the number of bus travel time under 68 seconds has increased 40% to 50% by implementing both strategies of green time extension and red time reduction when controllable time limit is set within 6 seconds. It is shown that the proposed signal preemption control strategies do significantly benefit the bus timetable adherence.
Although only three signalized intersections are included in the case study, it is believed the same control logics can be applied to other scenarios with more intersections or longer road segments. With the coming of ITS age, the GPS positioning information will become much more available in modern city bus systems. Therefore we are more confident that the performance of this integrated framework of the bus travel time prediction model and signal preemption control strategies will definitely improve our real world bus timetable adherence.

1.李俊賢，「在靜態模型中運用傅立葉轉換分析隨機性動態旅行時間之研究」，台灣大學土木工程學研究所交通工程組博士論文，1996。
2.李穎，「類神經網路應用於國道客運班車旅行時間預測模式之研究」，國立成功大學交通管理研究所碩士論文，民國90年。
3.何志宏等，「先進式微觀車流模擬器－PARAMICS 模擬軟體應用於高速公路與市區道路交控系統整合策略研究」期中報告，交通部運輸研究所，民國90 年7 月。
4.吳健生，「公車優先號誌之設計與應用」，都市交通81期，民國84年5月
5.吳淵展，「公車絕對、條件與補償優先號誌時制之研究」，中央土木工程研究所碩士論文，民國87年。
6.吳佳峰，「有GPS資訊提供下之車輛旅行時間預估模式之研究」，交通大學運輸工程與管理學系碩士班碩士論文，2001。
7.胡大瀛、洪百賢，「公車動態資訊應用設計競賽成果報告」，公車動態資訊應用設計競賽論文集，151-158頁，2002。
8.胡守任等，「公車到站時間預估模式建立與實作」，公車動態資訊應用設計競賽論文集，125-149頁，2002。
9.張鈞萍，「發展緊急救援車輛之行車路徑導引系統與號誌優先通行控制邏輯之研究」，國立成功大學交通管理研究所碩士論文，民國91年。
10.陳陸陣，「台鐵行車延誤補償制度之研究」，國立交通大學運輸工程與管理學系碩士論文，民國89年。
11.劉方旗，「市區公車排班與即時機動調度之研究-以新竹客運為例」，國立交通大學交通運輸研究所碩士論文，民國87年。
12.魏健宏等，「台北市專用道旅行時間預測之初步研究」，公元2000台灣智慧運輸系統ITS國際研討會暨展覽會，479-495頁，2000。
13.顏贊峰，「公車優先號誌控制策略於無設站路口之應用」，中央大學土木工程研究所論文，民國85年。
14.蘇志強、莊弼昌、李紹榆，「全觸動公車優先通行號誌控制模式之研究」，中華民國運輸學會第十三屆論文研討會，中華民國87年12月。
15.蘇志強，「公車優先通行與調適性交通號誌整合控制之研究」，警學叢刊第26卷2期，民國84年9月。
16.蘇志強，李政聰，王銘亨，「幹道公車優先通行續進號誌控制模式之研究」，第一屆台灣ITS 國際研討會論文集，1999 年04 月。
17.Abdelfattah, A. M. and Khan, A. M., “Models for predicting bus delays”, Transportation Research Record 1623, pp. 8-15, 1998.
18.Benvelli, D. A., Radwank, A. E., and Hurley, J. W., “Evaluation of Bus Preemption Strategy by Use of Computers Simulation”, Transportation Research Record, NO.906,1983, pp.60-67.
19.Bill Alan Cisco and Snehamay Khasnabis, “Technique to Assess Delay and Queue Length Consequences of Bus Preemption”, Transportation Research Record, No.1494, pp.167-175, 1995.
20.Bae, S. and Kachroo, P., “Proactive travel time predictions under interrupted flow condition”, Vehicle Navigation and Information Systems Conference, pp.179-189, 1995.
21.Chien, I J., Shin, C. H., and Park, I., “Dynamic Bus Arrival Time Artificial Neural Network”, Journal of Transportation Engineering, September 6, 2001.
22.Hunter-Zaworski, K. M., Kloos, W. C., and Danaher, A. R., “Bus Priority at Traffic Signals in Portland: The Powell Boulevard Pilot Project”, Transportation Research Record, No. 1503, 1995, pp. 29-33.
23.Huang, W. J., “Information service for buses operating on bus-lanes in Taipei”, Presented at the 9th World Congress on ITS, Chicago, Illinois, 2002.
24.Joseph A. Wattleworth, Kenneth G. Corage and Charles E. Wallace, “Evaluation of Bus-Priority Strategies on Northwest Seven Avenue in Miami”, Abridgment, pp32-35.
25.Jarjees, F. G., “Prediction of vehicle travel times and traffic conditions”, a doctoral dissertation of technology of Sydney, pp. 199-230, 2002.
26.Ludwick, J. S., “Bus Priority System: Simulation and Analysis”, Final Report, The Mitre Corporation , Prepared for the U.S. Department of Transportation. Report No.: UTMA-VA-06-0026-76-1, 1976.
27.Richardson, A. J. and Ogden, K. W., “Evaluation of Active Bus-priority Signals, Committee on Bus Transit System, 1979,pp.5-12。
28.Snehamay Khasnabis, Gangula V. Reddy and Syed Khurshidul, “Evaluation of Operating Cost Consequence of Signal Preemption as an IVHS Strategy”, Transportation Research Record, No.1390, pp.3-9, 1993
29.Shin, S. W. et al., “Randomness of de-noised time series data for travel time forecasting”, Presented at the 6th World Congress on ITS, Toronto, Canada, 1999.
30.Welding, P. I., “The instability of closed-interval service”, Operational Research Quarterly, Vol. 8, pp. 133-148, 1957.