Competition on Urban Traffic Signal Control Problems
Loading...
Authors
EL Abd, Mohammed
Gao, Kaizhou
Issue Date
2022
Type
Technical Report
Language
en_US
Keywords
Alternative Title
Abstract
According to the recent study in [1], one clear aspect that is lacking in the domain of Traffic Signal Control (TSC) is the availability of common libraries for benchmarking and comparing different algorithms. Metaheuristics have been employed in the domain of TSC [2-6]. A vast majority of publications apply proposed algorithms on their own generated artificial networks and/or networks from the authors’ cities/countries without comparing with previous work. This poses a huge challenge in identifying the state-of-the-art in TSC. A comprehensive literature review was presented in [1] in chronological order and rarely it was found that authors provide comparisons with existing works. Having benchmark libraries containing networks of different sizes and under different conditions (undersaturated, saturated, and oversaturated) would ensure that a proposed algorithm is properly evaluated and would provide the research community with a baseline against which algorithms could be fairly evaluated.