Modelling saturated traffic networks using massively parallel computing techniques

Lu, K.H.; Jones, J.; Kerridge, J.

Abstract

A technique for modelling the flow of motor vehicle traffic in congested urban signal-controlled networks using parallel computing is presented. A brief overview of traffic simulation and control precedes a discussion of the future requirements for controlling urban traffic. It is argued that a computer model which can represent the attributes and movement of every individual vehicle in the network in real time is a core requirement of any future control system. It is suggested that only parallel computing can provide the necessary computational power and if costs are to remain acceptable the transputer and its programming language, OCCAM, are an obvious solution. Both the transputer and OCCAM are described along with a brief introduction to the techniques and benefits of parallel computing and MOSTFLOW, a microscopic and parallel process-oriented traffic network model which is being developed to meet the above needs, is presented.

• Journal Article Type: Article
• Acceptance Date: Apr 1, 1994
• Publication Date: 1994
• Deposit Date: Apr 15, 2019
• Journal: Traffic engineering & control
• Print ISSN: 0041-0683
• Peer Reviewed: Peer Reviewed
• Volume: 35
• Issue: 7/8
• Pages: 433-436
• Keywords: Vehicle traffic, road network, mathematical model, computer programs.
• Public URL: http://researchrepository.napier.ac.uk/Output/1730088