基於行動通訊網路之觀光區道路即時交通資訊與旅行時間預測

Abstract

提供即時交通資訊能夠幫助用路人得知目前交通狀況，進而節省旅行時間及減少空氣油耗等問題，利用手機訊號作為定位的訊號，並且利用無線通訊網路追蹤使用者位置，透過多個手機即時訊號產生目標道路即時交通資訊，這樣的機制不需要花費巨大的費用來建置或維護傳統的車輛偵測裝置，這種裝置損壞率及建設費用都相當高昂，並且只能提供固定地點的交通資訊。相較之下，透過無線通訊網路訊號追蹤分析使用者的移動行為，可以提供更即時、全面的交通資訊，而且降低了額外設備費用。 本研究提出利用NLU與其他事件的組合推估行車時間，，藉由組合的時間差，可推估目前實際行車時間；NLU為強制性位置更新行為，當使用者移動至另一個LA便會產生NLU事件，我們追蹤這些使用者並依照事件發生時間的先後順序給予編號，當使用者在目標路段產生其他事件時並且藉由觀察路段出現的使用者評估道路上使用者數量(Vehicle Count Index, VCI) ，作為預測旅行時間的基準並結合歷史塞車時期的旅行時間，提供目前使用者其可能行車時間。

Providing real-time traffic information to road users can help them to save travel time and reduce air pollution. Since MSs register their locations with cellular networks constantly, the control signals of cellular network can be used to obtain real-time traffic information. This method has been referred to as Cellular Floating Vehicle Data (CFVD). CFVD is more cost-effective than vehicle detectors or GPS-equipped probe cars because it doesn’t need any additional on-road or on-vehicle devices. However, most CFVD researches focused on freeways and urban roads, but not on tourist areas located in suburban areas. In this thesis, we present a CFVD algorithm to estimate and predict the travel time of an arterial road in a tourist area. The entrance of the arterial road under study locates in the boundary of two location areas (LAs), The time difference between an NLU performed at the entrance and an event performed on the arterial road is used to estimate the travel time. In addition, the number of NLUs performed at the entrance is used to generate Vehicle Count Index (VCI) as an indicator of the number of road users on the monitored roads. We use this index and historic data to predict the travel time of the monitored road.