Our newest research results have been accepted for publication in four conference papers.
The first paper, titled “Proposal for Fuzzy based Adaptive Traffic Signal Control for Reducing Vehicle Emissions” is accepted for the conference “3rd International Scientific Conference “TRANSPORT FOR TODAY’S SOCIETY” which will be held on October 14-16, 2021 in Bitola, North Macedonia. This paper was made in the co-authorship of Daniela Koltovska Nechoska, Roberto Pashic, Robert Šimić, and Edouard Ivanjko. It presents a proposal on improving adaptive traffic signal control with emphasis on reducing road vehicle emissions using fuzzy logic reasoning and a newly developed traffic and emissions sensor.
The second paper, titled “Influence of Spatial Placement of Variable Speed Limit Zones on Urban Motorway Traffic Control”, is accepted for the “International Scientific Conference The Science and Development of Transport ZIRP2021” which will be held from September 30 until October 1, 2021, in Šibenik, Croatia. This paper was made in the co-authorship of Miloš Josip, Hršak Patrik, Topić Nikola, Jakšić Leon, Kušić Krešimir, Vrbanić Filip, and Ivanjko Edouard. It presents the results of an analysis regarding the placement of speed limit zone for variable speed limit control systems. Emphasis is on traffic parameters and emission parameters. This paper is also recommended for the scientific journal “PROMET – Traffic&Transportation” due to its high quality.
The third and fourth papers are accepted for presentation at the scientific conference“63rd International Symposium ELMAR-2021, Special session Intelligent Transport Systems”, which will be held on September 12-15 in Zadar, Croatia. The third paper titled “Creating Representative Urban Motorway\\ Traffic Scenarios: Initial Observations” was made in the co-authorship of Filip Vrbanić, Mladen Miletić, Edouard Ivanjko, and Zeljko Majstorović. It presents the first results of traffic data analysis regarding Slovenian motorways in order to obtain realistic simulation scenarios for testing learning-based traffic controllers.
The fourth paper, titled “Combining Neural Gas and Reinforcement Learning for Adaptive Traffic Signal Control,” was made in the co-authorship of Mladen Miletić, Edouard Ivanjko, Sadko Mandžuka, and Daniela Koltovska Nečoska. It presents the results of the development of a reinforcement learning-based adaptive traffic signal controller with fast convergence and good control quality. Its results are an improvement of the ELMAR-2020 paper titled “State Complexity Reduction in Reinforcement Learning-based Adaptive Traffic Signal Control” published last year.