交通指派方法的驗證與比較分析-以台北市為例

由於運輸管理策略實際應用時風險很大，因此通常會以建模的方式先進行績效評估。然而於交通指派前，須先針對運輸需求分析軟體進行路網及旅運資料校估以反映現況，本研究使用VISUM為工具，建立一個台北市的都市路網，進行供給及需求模型之校估。交通指派是運輸規劃中的一個重要步驟，它是將調查得到的起迄點之間的旅次分佈數據 (O-D矩陣，按照一定的規劃分配到路網中的各條道路上，從而推測各條道路上的交通量，並以此作為路網規劃及設計的依據。目前國際上較通用、較成熟的指派模型主要有:一般均衡模型、非均衡模型，而VISUM中整合的交通指派模型基本覆蓋以上幾種通用模型,本文運用VISUM對城市道路的交通模型模型進行研究。 VISUM軟體的校估流程及交通指派方法的績效評估，可供未來相關研究人員參考使用。本研究將以台北市為例，原因在於台北市區道路容量已無擴增之可能，故需提高台北市區的運輸管理；加之歷史資料庫建置完整，得以進行相關研究。本研究首先以VISUM進行建模，利用T-flowfuzzy方法更新O-D矩陣及修改道路參數，為使路網盡可能呈現實際道路狀況，再使用四種不同的交通指派方法得到各道路之交通流量，最後利用10個指標以篩選出較具優勢的方法，提供未來運輸規劃研究者一個具備效率的建模流程。

關鍵字

VISUM ；交通指派； O-D矩陣推算；路網績效評估

並列摘要

Due to the high risk of practical application of transportation management strategies, performance evaluation is usually performed in a modeling manner. However, before assigning traffic volume, we must first carry out road network and travel data evaluation for transport planning software to reflect the current situation. This study uses VISUM as a tool to establish a metropolitan road network in Taipei to conduct a supply and demand model assessment , and then use different assignment methods for overall comparison. Traffic assignment is an important step in transportation planning. It is the distribution data of the travel time between the starting and the return points (the OD matrix, which is allocated to each road in the road network according to a certain plan, so as to infer the various items. The volume of traffic on the road is used as the basis for road network planning and design. At present, the more general and mature assignment models in the world mainly include: general equilibrium model and non-equilibrium model, and the traffic assignment model integrated in VISUM basically covers The above several general models, this paper uses VISUM to study the traffic model model of urban roads. The VISUM software calibration process and performance evaluation of traffic assignment methods can be used by future relevant researchers. This study will take Taipei as an example. The reason is that the capacity of roads in Taipei City has not been expanded. Therefore, it is necessary to improve the transportation management of Taipei City. In addition, the completion of the historical database will enable relevant research. In this study, VISUM was first used to model the road network as much as possible, and the O-D matrix was updated using the T-flowfuzzy method. Secondly, compare four methods of traffic assignment and use 10 indicators to screen out more advantageous methods to provide future transportation planning researchers with an efficient modeling process.