大型事件起迄矩陣推估之研究－以澳門格蘭披治大賽車為例

Abstract

起迄矩陣為運輸規劃中相當重要的輸入值，但傳統旅運需求調查方法需耗費許多人力與成本才能取得完整資料，因此，需要其他成本較低且準確性高之方法來取得起迄矩陣。而常見的方法為透過路段觀測流量反向推估或修正實際的起迄矩陣，使模式流量結果與觀測值較為接近。而本研究在推估過程中嘗詴二種推估方式，同步推估與依序推估，並比較結果上之差異。 另外，本研究亦進行大型事件起迄矩陣推估，係因許多城市每年都有舉辦大型事件，活動期間易吸引大批民眾聚集，需有良好的交通管制措施來分散車流。而交通規劃與管理需要有效的交通需求數據支持分析的工作，但這些事件都只持續數天，資料收集更是不易。本研究以澳門每年於11月舉辦的格蘭披治大賽車為例，使用帄日的旅運需求調查，帄日的市區路段交通量調查，並於格蘭披治大賽車期間調查少量路段的交通量數據，推估賽事期間的起迄矩陣以及市區整體交通的狀態變化，期望結果可作為賽事期間交通控制和管理之參考依據。而整體研究結果顯示，若能透過起迄矩陣推估步驟，可更貼近實際的交通狀況，此亦代表模式能有效推測賽事期間當地居民的旅運起迄需求變化及整體路網車流狀況的改變。

The origin-destination (O-D) demand matrix is an important input in the planning of transportation. However, obtaining the demand information using the traditional household survey method is costly in money and time, and O-D estimation from link traffic counts is a common methodology to calibrate or update the O-D matrix to fit into the observations. This study also tests two different O-D estimation processes, which estimate motorcycles and private cars matrices simultaneously and in sequence. The differences of results are also compared. This study also estimates the matrices during the mass events. Mass events organized annually may attract a large amount of traffic and require a preplanned traffic control scheme, which requires reliable traffic demand data. Such data is usually not available as the event is usually too short for detailed data collection. In this study, the Macau Grand Prix event which is hold in streets in November is investigated. With the previously collected data from household survey and link traffic counts in the urban area, we propose a methodology to estimate the O-D demand and the corresponding changes in the city traffic pattern during the event with additional traffic count observations in a limited number of locations. The findings would be useful to the corresponding event organizers for traffic control and management purposes. The results of this study revealed that the traffic condition is similar to the real traffic pattern after O-D estimation procedure. And it represents the model can speculate about the change of local travel demand and whole traffic flows during the mass event.