使用階層式迴歸模型預估高速公路車流量

Abstract

對於那些要經常在尖峰時間使用高速公路的人而言, 避免不必要的擁擠以節省寶貴時間是相當重要的. 在本篇論文中, 我們將用一些統計的方法來分析高速公路的車流資料並且發展一種方法在高速公路各檢測器上預測短期的車流變化. 我們將使用考慮時間, 日期, 和上游車流量的hierarchical 迴歸模型. 我們用高速公路北部路段的車流量當作我們所要分析的資料.

For those who usually use Freeway at peak time period, predicting the car flows at these periods and avoiding unnecessary congestion in order to save time is important. In this thesis, we want to use statistical model to analyze the freeway flow data and develop a technique in predicting short-term flow changes on sequence of detectors in a freeway network. We use hierarchical regression model which considers time of day, day of week and upstream detector flows. The data we analyze come from Taiwan Area National Freeway Bureau.