高速鐵路運量預測模式重建與分析

High Speed Rail Ridership Forecast Study-Model Rebuilt and Analysis

林威任(Wei-Ren Lin)

指導教授 ： 許聿廷

國立臺灣大學/工學院/土木工程學研究所/碩士(2014年)

https://doi.org/10.6342/NTU.2014.01899

摘要

高速鐵路具備高速度、高運能及低污染等優點，是相當具有吸引力之城際運具。自2007年台灣高鐵通車以來，即大幅縮短台灣西部走廊各大城市間的旅行時間，成就「一日生活圈」之概念。由於建設高速鐵路需要相當龐大的投資成本，若由政府主持興建，受到行政程序繁複即財政負擔的影響，將可能使通車時程不斷延宕。近年許多國家將大型公共建設轉向私有化的替代方案。台灣高鐵是世界上最大的BOT建設案，但自通車以來，營運即受到運量不足的影響，造成龐大的財務赤字。於2009年因財務困難，迫使政府不得不提供資金援助。1990年代，交通部高鐵籌備處所委託幾家顧問公司預測高鐵的運量，但大多報告皆樂觀預估運量及營收，是造成後續財務規劃不佳的成因之一。然而由於實際通車日期與當初預測時程不同，許多社會經濟條件已不可同日而語，在20年後的今日與當初所預期之條件大多不符，如GDP、人口、基本費率及行車時間皆不同。此外，台灣社會經濟環境和人文地理因素與其他國家有很大的不同，委託之顧問公司多以他們的經驗來推測需求量，但未必適用於台灣。本研究欲根據台灣高鐵的經驗，重新建構高速鐵路運量預測模式。
該模式分為兩個部分，一為直接總體需求模式，另一為總體羅吉特模式（運具分配模式）。利用歷史巨觀資料校估模式參數，並以台灣高鐵通車後之實際數據驗證模式參數。與實際運量相比，本研究所提出之高鐵運量模式預測誤差約為15%。以次模式預測20年內的高鐵運量及營收值，結果顯示，即將營運通車的高鐵新增三站(苗栗、彰化及雲林)，在民國110年的預測情境下，可增加16％運量量及11％的營收，即對營運有正面幫助。但這三站所貢獻的運量及營收將比現有高鐵嘉義站更低。在高鐵票價策略方面，本研究結果顯示現今台灣高鐵所制定之票價可幾乎創造最大之營收，因此目前票價措施算是合適。在高鐵接駁時間方面，當接駁時間越少，可有效吸引短途旅客(尤其是通勤旅客)，若高鐵站與目前傳統鐵路車站共構，可減少大量的接駁時間，民國110年計算結果顯示會約有55％的運量提升及30％的營收增加。

關鍵字

台灣高鐵 ； BOT ； 運量預測 ； 總體資料

並列摘要

With its virtues of high speed, large capacity, reduced levels of energy consumption, and pollution, high-speed rail (HSR) is emerging as an attractive transportation system. Due to the large investment burden required for HSR projects and the inefficiency of government-sponsored public construction projects, many countries are now turning to the alternative of privatizing their HSR projects. However, the private company often encounters financial deficit when the operating revenue cannot balance the cost. Taiwan high speed rail (THSR) is the largest BOT (Build-Operate-Transfer) case in the world. It once declared bankrupt because of financial difficulties until the government provided funding support in 2009. Some criticisms focused on the overestimated ridership forecasts made by several consultants and institutes. However, the actual operation period is different from the planned one, which involves the changes of socioeconomic, environmental, and geographic backgrounds (such as GDP, population, and fare price). Also, the employed demand forecast models may be characterized by the experiences in other countries, which can be inapplicable in Taiwan. Hence, this study seeks to rebuild HSR ridership forecast model based on the experiences of THSR and re-examine it using actual operation data.
This model is composed of two parts, direct demand forecast model (overall inter-city travel demand), and aggregate logit model (mode split model). The parameters in both models are calibrated by using historical aggregate data and verified by real ridership data after THSR opening, showing that the prediction error of the proposed models is within 15%. The current fare price of THSR can reach the nearly best revenue, implying appropriate pricing strategies. The prediction of THSR ridership and revenue for the next 20 years are conducted based on the developed models. The results show that the three stations (Miaoli, Changhua, and Yunlin) which are currently under construction may increase 16% ridership and 11% revenue in 2021, highlighting their positive effects on future operation. However, the demand generated from each of these three stations may be lower than the one currently with the lowest demand (Chiayi). If THSR stations were built on the conventional railway stations, access time can be significantly reduced, which may lead to 55% ridership and 30% revenue increase. The reduced access time may also attract more short-distance travelers (commute trips).

並列關鍵字

Taiwan high speed rail ； BOT ； ridership forecast ； aggregate data

參考文獻

47.凌瑞賢(2004)，「運輸規劃原理與實務」，鼎漢國際工程顧問股份有限公司。

1.Ashiabor, S., H. Baik, and A. Trani. Logit models for forecasting nationwide intercity travel demand in the United States.Transportation Research Record, No. 2007, 2007, pp. 1–12.

3.Chen, Z.H., (2010). Analysis of factors affect high speed train ridership in the United States -The acela express case study, 51st Annual Transportation Research Forum 2010, Vol. 1, pp. 1-15.

5.Goolsby, M.E. and Walker, E.L., (1993). Texas High Speed Rail ridership forecasting, High Speed Ground Transportation Systems I, pp. 256-265.

6.Peters, J.C., Han, E-P., Peeta, S. and DeLaurentis, D., (2013), Analyzing the Potential for High-speed Rail as Part of the Multimodal Transportation System in the Midwest Corridor.