01698237

Component

Traffic planning and operation calls for network modes with different vehicle classifications. In real traffic flow, cars and trucks have different attributes including free-flow speed, capacities, critical and jam densities, and thus asymmetric affects on each other while propagating in the networks. Recent years have also witnessed tremendous traffic data by cars and trucks. How to make best use of those data for multi-class network models is unclear. In this research, the authors build and implement a large-scale mesoscopic dynamic traffic assignment model considering separate demand and flow dynamics for cars and trucks. They incorporate heterogeneous traffic characteristics into the cell transmission model and the link queue model. The authors' mesoscopic car-truck network model can be efficiently calibrated using multi-class traffic data, such as counts and speed, and it is designed for efficiently evaluating traffic planning and operation strategies on large-scale regional traffic networks. Through a real-world case study in Pittsburgh region, the authors show the data-driven car-truck model fits spatio-temporal data reasonably well, and performs effectively to analyze the impact of a local development project on a seven-counties regional traffic network.

This paper was sponsored by TRB committee ADB30 Standing Committee on Transportation Network Modeling. Alternate title: Data-driven Mesoscopic Network Modeling With Cars and Trucks: A Case Study in Pittsburgh.

19-02321

English

6p

2019

Transportation Research Board 98th Annual Meeting

Digital/other

Figures; References; Tables

Automobiles; Case studies; Dynamic traffic assignment; Mesoscopic traffic flow; Traffic simulation; Trucks

Cell transmission model

Pittsburgh (Pennsylvania)

Highways; Planning and Forecasting; Vehicles and Equipment

Transportation Research Board Annual Meeting 2019 Paper #19-02321

TRIS, TRB, ATRI

Dec 7 2018 9:49AM