Étude des instabilités dans les modèles de trafic

Abstract : Highway traffic is known to be unstable when the vehicle density becomes too high, and to create stop-and-go waves, with an alternance of free flow and congested traffic. First-order traffic models can't reproduce these oscillations, but higher-order models can, both microscopic (car-following models) and macroscopic (systems of conservation laws).This thesis analyses the representation of unstable traffic states and oscillations in various traffic models. At the microscopic level, because of the flux concavity, the average flow of these oscillations is lower than the equilibrium flow for the same density. An algorithm is given to stabilize the flow with multi-anticipation, using an intelligent autonomous vehicle.At the macroscopic level, this work introduces averaged models, using the fact that the spatio-temporal scale of the oscillations is too small to be correctly predicted by simulations. The averaged LWR model, which consists of two conservation laws, enables a macroscopic representation of the density variance in a heterogeneous traffic, and gives the correct average flow of these states. A comparison with the ARZ model, also of order 2, shows that the averaged model can reproduce a capacity drop in a more realistic way.Finally, this thesis presents the SimulaClaire project of real-time traffic prediction on the ring road of Toulouse, and its parallelized parameter optimization algorithm
Document type :
Theses
Complete list of metadatas

Cited literature [53 references]  Display  Hide  Download

https://pastel.archives-ouvertes.fr/tel-01526684
Contributor : Abes Star <>
Submitted on : Tuesday, May 23, 2017 - 1:28:35 PM
Last modification on : Friday, July 7, 2017 - 1:17:55 PM
Long-term archiving on : Friday, August 25, 2017 - 12:39:01 AM

File

TH2016PESC1067_convertie.pdf
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-01526684, version 1

Collections

Citation

Rémi Sainct. Étude des instabilités dans les modèles de trafic. Mathématiques générales [math.GM]. Université Paris-Est, 2016. Français. ⟨NNT : 2016PESC1067⟩. ⟨tel-01526684⟩

Share

Metrics

Record views

299

Files downloads

250