應用條件逆矩陣方法推估依時性起迄對矩陣之研究

A Conditional Inverse Approach for Inferring Time-Dependent Origin-Destination Matrices

王中允(Chung Y. Wang) ； 吳玫茵(Mei Y. Wu)

《國防管理學術暨實務研討會》 第十九屆 (2011/06) Pp. 616-638

https://doi.org/10.29497/GFGL.201106.0616

摘要

在運輸學術領域中，利用路段流量反推起迄對旅次矩陣，是近二、三十年來研究重點。隨著車輛偵測器直接觀測道路車流狀況的廣泛應用，讓利用路段流量反推起迄對旅次矩陣更具實務上研究的潛力。本研究利用條件逆矩陣反推起迄對旅次矩陣的求解程序，應用於推估依時性的起迄對旅次矩陣中。研究證明此求解程序能準確的反推出依時性的起迄對旅次矩陣，且能同時應用於使用者均衡及非使用者均衡條件下的路網，較傳統的反推起迄對旅次矩陣的模型更具準確性及一般性。

關鍵字

依時性 ； 起迄對旅次矩陣 ； 條件逆矩陣

並列摘要

In general, the link traffic flows of network are observed which resource from each trip original area and go to each trip destination area. That is, by a suitable estimation method, the origin-destination (O-D) trip matrices may be inferred from link traffic flows. Especially in recent years, the vehicle detectors (VD) are used wildly. The data of link traffic flows can be obtained easily through VDs. In this study, the conditional inverse approach is adopted to estimate time-dependent O-D trip demands through a linear equation system which included the flow conservation relationship between link flows, path flows and O-D trips. Through testing, use the solution procedure in different networks is inferring results accurately. The present study is a more general estimation time dependent O-D demand trips method and more application potential in transportation practice operation.

並列關鍵字

Time-Dependent ； Origin-Destination trip Matrices ； Conditional Inverse