An investigation of proportionally fair ramp metering

Add to your list(s) Download to your calendar using vCal

• R. J. Gibbens, Computer Laboratory, University of Cambridge
• Monday 26 April 2010, 15:00-16:00
• Seminar Room 1, Isaac Newton Institute for Mathematical Sciences.

If you have a question about this talk, please contact Neil Walton.

This talk concerns ramp metering which is one important approach to dealing with congestion on motorways. Congestion occurs when demand exceeds available resources and can significantly reduce the capacity of the motorway network at peak times. Reduced capacity results in additional delays, increased environmental pollution and hinders passenger safety. Congestion is observed to cause low but highly volatile speeds resulting in more uncertain journey times (referred to as flow breakdown or stop-and-go behaviour).

Ramp metering is intended to control the entry of new flow in such a way as to maintain steady flow on the motorway and to avoid the flow breakdown associated with congestion. The rate of entry of flow is set according to the particular ramp metering strategy. Such strategies have been the subject of much attention in the transport literature. One of the key issues is the trade-off between efficiency and fair use of resources. This is a trade-off that has been considered extensively in the modelling and control of communication networks.

This talk adds to recent work (Kelly & Williams (2010)) on a ramp metering strategy, proportionally fair metering, inspired by rate control mechanisms developed for the Internet. Specifically, we use simulation results to compare proportionally fair metering with a greedy strategy for a linear network with a series of entry points leading towards a single common destination for all the traffic, such as a radial route towards a city centre. Under our modelling assumptions, the greedy strategy is provably optimal for exogenously determined arrival streams of traffic, but it is unfair, in a certain precise sense, between different entry points and may well have perverse and suboptimal consequences if it influences traffic demand. We further consider a network with parallel roads where flows of traffic may have route choice according to the levels of queueing at the individual entry points.

JOINT WORK WITH F .P. KELLY

This talk is part of the Optimization and Incentives Seminar series.

This talk is included in these lists:

• All CMS events
• All Talks (aka the CURE list)
• CMS Events
• Cambridge talks
• DPMMS Lists
• DPMMS info aggregator
• DPMMS lists
• Economics and Computer Science Talks
• Hanchen DaDaDash
• Interested Talks
• Optimization and Incentives Seminar
• School of Physical Sciences
• Seminar Room 1, Isaac Newton Institute for Mathematical Sciences
• Statistical Laboratory info aggregator
• Trust & Technology Initiative - interesting events
• bld31

Note that ex-directory lists are not shown.