01369994

Component

The equivalent mathematical formulation of the combined doubly-constrained gravity-based trip distribution and paired-combinatorial-logit stochastic user equilibrium assignment problem (CDA-PCL-SUE) is proposed. Its first order conditions are shown to be equal to the gravity equations and PCL formula. The proposed solution method is a disaggregate simplicial decomposition algorithm that iterates two phases. Phase I employs the partial linearization descent algorithm to approximately solve the restricted CDA-PCL-SUE, and Phase II is the column generation phase. In Phase I, the partially linearized problem is decomposed into an entropy maximization problem on O-D flow space that can be solved by Bregman’s balancing algorithm and a PCL SUE problem that can be solved by PCL formula. CDA-PCL-SUE is compared with its multinomial-logit counterpart (CDA-MNL-SUE) on two test networks at two congestion levels. The authors found that the relationship of O-D flow difference and dispersion factor is not necessarily clear, whereas the link flow patterns from the two models are more identical at higher dispersion factors on the two networks at two congestion levels. At a fixed dispersion factor, CDA-PCL-SUE assigns less flow to a path with a higher average similarity index and higher path cost than CDA-MNL-SUE does. CDA-PCL-SUE generally assigns less flow to links with more paths passing through than CDA-MNL-SUE does. The relationship between O-D flow allocation and the average similarity indices for O-D pairs is not obvious, since the gravity-based trip distribution equation is based on additional variables and the origin and destination demand constraints must be satisfied.

This paper was sponsored by TRB committee ADB40 Transportation Demand Forecasting

01362476

12-0591

English

31p

2012

Transportation Research Board 91st Annual Meeting

Digital/other

Figures; References; Tables

Algorithms; Combinatorial analysis; Logits; Origin and destination; Stochastic processes; Traffic equilibrium; Trip distribution

Highways; Planning and Forecasting; I71: Traffic Theory

Transportation Research Board Annual Meeting 2012 Paper #12-0591

TRIS, TRB

Feb 8 2012 4:55PM