Abstract
Urban planning policies are increasingly questioning the problem of the last kilometer. The objective of this demonstration is to show how, with the help of a multi-agent simulation, it is possible to measure the effectiveness of different management strategies for traffic lights. Based on 2 types of environments and 5 different strategies, we show how to evaluate a road infrastructure. The software architecture presented is easily extensible to any kind of traffic light or car behaviors and any map. Our demonstration is carried out with the SUMO simulator with the help of both abstract road maps and real maps from different cities (parts of Nice, Manhattan, Madrid).
Author supplied keywords
Cite
CITATION STYLE
Mathieu, P., Nongaillard, A., & Thery, A. (2020). A Multi-agent Evaluation of Traffic Light Models. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 12092 LNAI, pp. 451–455). Springer. https://doi.org/10.1007/978-3-030-49778-1_46
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
