A probabilistic model for the origin of motorway shockwaves
A probabilistic model for the origin of motorway shockwaves
From stop-start driving through to stationary queues, traffic congestion in its various forms is becoming an increasingly common feature on the motorway network. Flow-breakdown on motorways is a complex phenomenon that occurs under a combination of specific traffic conditions and perturbations in the flow.
The context of this thesis is the value of being able to pre-empt the upstream propagation of the sequential braking pattern (known as a shockwave) through dense traffic, thereby providing an opportunity to influence the traffic using speed control. The current body of literature describes flow-breakdown in various forms, including car-following behaviour, merge conflicts, gap acceptance, and fluid models for shockwave propagation. However, the area of shockwave origins (seed-points) is relatively unexamined.
A fresh representation of shockwaves was obtained from traffic data on the M25 London Orbital motorway, and this led to a deeper investigation into the properties and mechanisms behind the formation of shockwaves. This investigation comprised a six-month survey of traffic data (accompanied by video footage) in order to examine the variety and distribution of events that initiate shockwaves.
The implementation of automatic variable speed-limits on the motorway network raises the possibility for dynamic shockwave damping. A simple feasibility model is presented in this thesis, showing the importance of being able to predict the start of a shockwave in order to provide adequate time to influence drivers upstream of the origin point.
Research then focussed on how to model the origin of the shockwave, with a primary emphasis on statistical inference. A probabilistic model for the origin of motorway shockwaves is presented, requiring a detailed description of conditions present at the commencement of a shockwave. Detection of seed-points was therefore critical to the implementation of this probabilistic model.
Examination of existing methods of incident detection within the data set showed that whilst they detected the body of a shockwave effectively they did not extend to finding the origin point. Software has therefore been developed to identify seed-points, and descriptions of the proposed algorithms used in the software are presented and assessed on their ability to identify the origin of shockwaves. This study has shown that it is possible to identify the location of seed-points by classification using traffic characteristics. The primary variables used were averaged speed and occupancy (a proxy for headway) as obtained from MIDAS detectors on the M25 Controlled Motorway.
University of Southampton
Abou-Rahme, Nabil F
f40684ff-c763-4215-bbf5-aff44d14daee
2003
Abou-Rahme, Nabil F
f40684ff-c763-4215-bbf5-aff44d14daee
Abou-Rahme, Nabil F
(2003)
A probabilistic model for the origin of motorway shockwaves.
University of Southampton, Doctoral Thesis.
Record type:
Thesis
(Doctoral)
Abstract
From stop-start driving through to stationary queues, traffic congestion in its various forms is becoming an increasingly common feature on the motorway network. Flow-breakdown on motorways is a complex phenomenon that occurs under a combination of specific traffic conditions and perturbations in the flow.
The context of this thesis is the value of being able to pre-empt the upstream propagation of the sequential braking pattern (known as a shockwave) through dense traffic, thereby providing an opportunity to influence the traffic using speed control. The current body of literature describes flow-breakdown in various forms, including car-following behaviour, merge conflicts, gap acceptance, and fluid models for shockwave propagation. However, the area of shockwave origins (seed-points) is relatively unexamined.
A fresh representation of shockwaves was obtained from traffic data on the M25 London Orbital motorway, and this led to a deeper investigation into the properties and mechanisms behind the formation of shockwaves. This investigation comprised a six-month survey of traffic data (accompanied by video footage) in order to examine the variety and distribution of events that initiate shockwaves.
The implementation of automatic variable speed-limits on the motorway network raises the possibility for dynamic shockwave damping. A simple feasibility model is presented in this thesis, showing the importance of being able to predict the start of a shockwave in order to provide adequate time to influence drivers upstream of the origin point.
Research then focussed on how to model the origin of the shockwave, with a primary emphasis on statistical inference. A probabilistic model for the origin of motorway shockwaves is presented, requiring a detailed description of conditions present at the commencement of a shockwave. Detection of seed-points was therefore critical to the implementation of this probabilistic model.
Examination of existing methods of incident detection within the data set showed that whilst they detected the body of a shockwave effectively they did not extend to finding the origin point. Software has therefore been developed to identify seed-points, and descriptions of the proposed algorithms used in the software are presented and assessed on their ability to identify the origin of shockwaves. This study has shown that it is possible to identify the location of seed-points by classification using traffic characteristics. The primary variables used were averaged speed and occupancy (a proxy for headway) as obtained from MIDAS detectors on the M25 Controlled Motorway.
Text
922699.pdf
- Version of Record
More information
Published date: 2003
Identifiers
Local EPrints ID: 465179
URI: http://eprints.soton.ac.uk/id/eprint/465179
PURE UUID: 758ac332-863c-4343-89a9-c27cdab71b44
Catalogue record
Date deposited: 05 Jul 2022 00:27
Last modified: 16 Mar 2024 20:00
Export record
Contributors
Author:
Nabil F Abou-Rahme
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics