台鐵旅運需求與預測之研究

• Main Author: 陳欣欣
• Other Authors: 林佐鼎
• Format: Others
• Language: zh-TW
• Published: 2001
• Online Access: http://ndltd.ncl.edu.tw/handle/75221580520162966276

碩士 === 國立成功大學 === 交通管理學系 === 89 === Any planning of railroad engineering and operation project is deeply relied on effective and accurate passengers travel demand forecasting. Most previous studies on railway travel demand analysis focused on the volume of passengers, management strategies, and passengers travel demand analysis with aggregate data. This study tries to establish passengers travel demand forecasting models based on a single station. Because passengers travel demand is always affected by the characteristics of season, week and holiday, the time series models using ticket sales data are applied, and travel demand from Kaohsiung station to other stations is analyzed. The empirical results are as fallows: 1. Traveling distance is an important factor that affects the trend of travel demand. 2. When the policy of two days'' holiday for each two weeks is implemented, the weekend volume of long distance is transfered from Saturday to Friday. 3. Long distance passengers travel demand shows the characteristics of three-month circle in monthly passengers travel demand model and seven-day circle in daily passengers travel demand model. 4. The total mean percentage error (MPE) from Kaohsiung to Taipei is -9.12％ and is —31.98％ from Kaohsiung to Taichung. The forecasting ability of northbound is better than that of southbound.