Dynamic traffic assignment in degradable networks: paradoxes and formulations with stochastic link transmission model

ABSTRACT

This paper proposes a simultaneous route and departure time choice (SRDTC) problem with fixed demand in a degradable transport network. In this network, travelers face with stochastic travel times. Their selection of routes and departure times follows the UE principle in terms of the mean generalized route cost, which is defined as the probabilistic dynamic user optimal (PDUO) principle. The proposed PDUO-SRDTC problem is formulated as a variational inequality (VI) problem. As a special case of PDUO-SRDTC problem, the PDUO route choice (PDUO-RC) problem is also proposed and formulated as a VI problem. Network degradation is defined on the degradation of the outflow capacity of each link. A Monte Carlo-based stochastic link transmission model (MC-SLTM) is developed to capture the effect of physical queues and the random evolution of traffic states during flow propagation to estimate mean generalized route costs. Both the extragradient algorithm and the route-swapping method with a variable sample size scheme are developed to solve the proposed VI problems. Numerical examples are developed to illustrate the paradoxical phenomena of the problems and the effectiveness of the solution methods. Numerical results show that constructing a new road, enhancing link outflow capacity, or reducing outflow capacity degradation can lead to poor network performance and it is important to consider both network degradation and queue spillback when making transportation policies aimed at improving network performance. The results also demonstrate that the variable sample size scheme can give a quicker and better solution than the traditional fixed sample size scheme.

KEYWORDS:

• Dynamic traffic assignment
• degradable network
• probabilistic dynamic user optimal
• stochastic link transmission model
• paradoxes

Acknowledgements

The authors are grateful to the two reviewers for their constructive comments.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was jointly supported by the National Natural Science Foundation of China [71371026, 71431003, 71522001], a grant from the Research Grants Council of the Hong Kong Special Administrative Region, China (HKU 17207214), and the Fundamental Research Funds for the Central Universities (JZ2016HGPB0736).