A Discrete-State Mean-Field Model for Dynamic Transportation Network Flow and Routing with Link Degradation and Recovery

2026-99-2432

11/20/2026

Authors
Abstract
Content
This paper studies a dynamic transportation network model with link degradation and recovery. Each link has a traffic state and a health state. The model uses a discrete-state mean-field formulation to describe the average evolution of link health in the network. The service capacity depends on the health state, while the degradation and recovery probabilities depend on the congestion burden. Therefore, traffic flow and link health affect each other over time. We establish several theoretical results. We show that the traffic and health states stay in their valid ranges, derive the update equation for the mean health state, and show that heavier congestion leads to worse health change. We also prove that, under finite-duration demand, the total number of vehicles in the network decreases after demand stops, and all links finally recover to the healthy state. Numerical tests on a toy network and the Berlin-Tiergarten network show clear degradation and recovery patterns, as well as different vulnerability levels across links. In the healthy-limit case, the results also show small flow and Total System Travel Time gaps compared with the user-equilibrium benchmark on the Berlin-Tiergarten network. This shows that the proposed model is consistent with user equilibrium when link degradation is turned off.
Meta TagsDetails
DOI
https://doi.org/10.4271/2026-99-2432
Citation
Yang, P. and Zhao, T., "A Discrete-State Mean-Field Model for Dynamic Transportation Network Flow and Routing with Link Degradation and Recovery," 2026 2nd International Conference on Intelligent Transportation and Future Mobility (ITFM2026), Xi'An, China, March 27, 2026, https://doi.org/10.4271/2026-99-2432.
Additional Details
Publisher
Published
10 hours ago
Product Code
2026-99-2432
Content Type
Technical Paper
Language
English