摘要: |
Traffic assignment is an essential and fundamental step in the transportation
planning and management processes (Sheffi, 1985; Patriksson, 1994; Bell and Iida,
1997). Given constant travel demands between each origin-destination (O-D) pair
(i.e., travelers), and travel cost functions for each link of the network (i.e.,
transportation network), the traffic assignment problem is to determine the traffic
flow pattern as well as network performance measures (e.g., total system travel
time, vehicle miles of travel, vehicle hours of travel, fuel consumption and
emission, etc.). In practice, most traffic assignment models are single user class
and make a number of modeling assumptions including: separability assumption on
the link travel time function (i.e., no interactions), deterministic user equilibrium
(DUE) or stochastic user equilibrium (SUE) without accounting for route
overlapping, and no side constraints to describe the limited supply of certain scarce
resources (e.g., link capacities) in a network which are shared by multiple vehicle
types (e.g., passenger cars and multiple truck types) or to limit certain classes of
vehicles (e.g., trucks) on underpasses due to height restriction, bridges due to
weight restriction, and prohibited lanes due to lane restriction. However, as truck
traffic continues to grow as a result of increasing freight shipments transported by
trucks, there is an increasing interest to model multiple vehicle classes separately,
especially in addressing the impacts of truck traffic on congestion, infrastructure
deterioration, safety, and environmental concerns in many urban cities. According
to the Bureau of Transportation Statistics (BTS), freight shipments transported by
trucks account for 71 percent by value in U.S. dollars and 76 percent by weight in
tons of all commodity shipments (BTS, 2014). Hence, the purpose of this proposal
is to develop advanced traffic assignment method and computation algorithm for
addressing the asymmetric vehicle interactions, route overlapping, and traffic
restraints in multi-class traffic assignment problems involving multiple types of
trucks. |