以票證紀錄資料研發鐵道旅客流分析系統

Development of Rail Passenger Flow Analysis System Based on Ticket Transaction and Gate Data

黃仲瑋(Chung-Wei Huang)

指導教授 ： 賴勇成

國立臺灣大學/工學院/土木工程學研究所/碩士(2015年)

https://doi.org/10.6342/NTU.2015.10753

摘要

對於鐵路營運者而言，掌握旅客旅次分布與系統營運狀況在時間與空間中的變化是一項重要的工作，鐵路營運者可藉此調整其服務策略並提升整體系統之服務品質。隨著自動收費系統在近年來的快速發展，使用系統紀錄之票證資料分析旅次與系統效益是最直接且迅速的方式。本研究希冀透過有限的票證資料，重建旅客在系統內的行為，進而推估並分析旅客旅次分布與系統使用效率，以作為後續服務及營運規劃、營運策略調整的依據。
本研究以臺鐵局之旅客票證資料為基礎建立旅客流分析系統。旅客流分析系統主要包含「旅客流指派模組」與「旅客流分析模組」，旅客流指派模組針對不同資料類型的票種逐一建立推估模式，推估旅客於系統中的流動情形。本研究提出兩種旅客指派模式，分別以不同的指派方法將旅客指派至列車上，得到旅客完整的流動情形。推估結果經過整合後可得到整體旅客於系統中的流動情形，而進一步將系統的容量資料如車站月台容量、列車載客容量等等加入計算，藉由旅客流分析模組可得知系統的使用情形以及列車乘載率等營運績效。
本研究針對臺鐵基隆至新竹區間下午尖峰一日之旅客旅次分布以及系統使用情形進行分析。旅客多集中在南下臺北至中壢區間，其中臺北車站於尖峰時間之空間使用率高達的90%，而列車乘載率在板橋至樹林區間最高可達到接近70%。本研究也以分析結果針對車站與列車的使用情形提出改善措施，輔以證明研究成果可實際反映出各時空下旅客之旅次分布以及系統之使用情形。

關鍵字

鐵路運輸 ； 旅客流 ； 自動收費系統 ； 旅客選擇行為

並列摘要

Capturing the dynamics in passenger trips and system utilization over time and space is extremely important for railway operators. These key information can allow the adjustment of service strategies and plan, as well as enhance the performance and utilization of the system. Taiwan Railways Administration (TRA) is seeking an approach to automatically estimate passenger trips and system utilization based on available ticket transaction and gate data. Previous studies have estimated passenger flow using automatic fare collection data, but these applications have been limited to a single stopping pattern and/or a single type of ticket (usually the smartcard). The TRA provides five types of train services with a number of different stopping patterns through four types of tickets. Each type of tickets has specific characteristics and record in terms of transaction and gate data. Consequently, using ticket transaction and gate data to determine passenger flow and evaluate system utilization is a highly complex task. In this research, we develop a comprehensive process and evaluation framework to map passenger flow and evaluate system utilization. Two assignment models for matching passenger to train service are proposed. Empirical case study indicates that the developed tool can successfully match passengers with train services to present trip distribution, section utilization, station utilization, and train utilization over time and space. The validation result shows that the probability based assignment model is more suitable for TRA. Moreover, possible service adjustments have been examined to demonstrate the potential impact. These applications demonstrate that the developed tool is not only capable of analyzing trips and utilization based on historical data, but is also useful as an evaluation tool for the what–if analyses of potential service adjustments.

並列關鍵字

Rail Transportation ； Passenger Flow ； Automatic Fare Collection System ； Passenger Choice Behavior

參考文獻

Asakura, Y., Iryo, T., Nakajima, Y. and Kusakabe, T. 2012. Estimation of behavioural change of railway passengers using smart card data. Public Transport 4(1): 1-16.

Asakura, Y., Iryo, T., Nakajima, Y., Kusakabe, T., Takagi, Y. and Kashiwadani, M. 2008. Behavioural analysis of railway passengers using smart card data. Urban Transport XIV: Urban Transport and the Environment in the 21st Century.

Attanucci, J. P. and Wilson, N. H. 2011. Bus passenger origin-destination estimation and related analyses using automated data collection systems. Journal of Public Transportation 14(4): 131.

Barry, J., Freimer, R. and Slavin, H. 2009. Use of entry-only automatic fare collection data to estimate linked transit trips in New York City. Transportation Research Record: Journal of the Transportation Research Board 2112: 53-61.

Barry, J., Newhouser, R., Rahbee, A. and Sayeda, S. 2002. Origin and destination estimation in New York City with automated fare system data. Transportation Research Record: Journal of the Transportation Research Board 1817: 183-187.