Addressing policy objectives of traffic control using evolutionary algorithms

Sha'Aban, J.; Tomlinson, A.; HEYDECKER, B.; Bull, Larry

Abstract

This paper presents preliminary results from an ongoing research project that is investigating traffic management and signal control of evolutionary algorithms that use machine learning. It describes the application of novel technologies of machine learning to develop a responsive traffic signal control system. The objective of this project is to investigate the use of evolutionary algorithms to provide a fully distributed responsive traffic signal control system. This can then be used to implement traffic management objectives as required, and to take advantage of a range of detection technologies. The objectives are to achieve competitive results in comparison to present traffic signal control policies for their specific objectives and to be able to use evolutionary algorithms successfully in traffic signal control. Details of the general approach and the results of experiments are presented.

Citation

Sha'Aban, J., Tomlinson, A., HEYDECKER, B., & Bull, L. (2002, September). Addressing policy objectives of traffic control using evolutionary algorithms. Paper presented at European Transport Conference 2002, Cambridge, UK

• Presentation Conference Type: Conference Paper (unpublished)
• Conference Name: European Transport Conference 2002
• Conference Location: Cambridge, UK
• Start Date: Sep 9, 2002
• End Date: Sep 11, 2002
• Publication Date: Jan 1, 2002
• Journal: Association for European Transport
• Peer Reviewed: Peer Reviewed
• Keywords: algorithms, evolution, learning, policy analysis, signals, simulation, traffic control, traffic signal control systems
• Public URL: https://uwe-repository.worktribe.com/output/1083922
• Additional Information: Title of Conference or Conference Proceedings : European Transport Conference 2002