Upravljanje promjenjivim ograničenjem brzine primjenom ojačanog učenja za prometne tokove s umreženim autonomnim vozilima

master's thesis

• Title (english): Reinforced Learning based Variable Speed Limit Control for Traffic Flows with Connected Autonomous Vehicles
• Author: Filip Strikinac
• Mentor: Edouard Ivanjko (mentor)
• Mentor: Filip Vrbanić (komentor)
• Committee member: Miroslav Vujić (predsjednik povjerenstva)
• Committee member: Martin Gregurić (član povjerenstva)
• Committee member: Sadko Mandžuka (član povjerenstva)
• Granter: University of Zagreb; Faculty of Transport and Traffic Sciences; (Division of Intelligent Transport Systems and Logistics); (Department of Intelligent Transport Systems); Zagreb
• Defense date and country: 2022-09-22, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Traffic and Transport Technology; Intelligent Transport Systems and Logistics

Abstract

Današnje autoceste dizajnirane su za pružanje većeg maksimalnog prometnog kapaciteta, ali se javlja sve veći problem premalog kapaciteta istih. Međutim, gradske autoceste se najčešće ne mogu širiti, stoga je potrebno tražiti druga rješenja kako bolje iskoristiti njihov raspoloživ kapacitet. Jedan od sustava iz domene inteligentnih transportnih sustava, koji se može iskoristiti u tu svrhu, je sustav za upravljanje promjenjivim ograničenjem brzine (eng. Variable Speed Limit Control - VSLC). Autonomna vozila (eng. Autonomous vehicle - AV) i umrežena autonomna vozila (eng. Connected Autonomous Vehicles – CAV) uvelike se testiraju i brzo razvijaju u posljednjih nekoliko godina. Da bi se ispitali potencijalni učinci mješoviti tokovi prometa moraju se simulirati u različitim prometnim scenarijima. U ovom diplomskom radu napravljen je QL-DVSL (eng. Q-Learning – Dual Variable Speed Limit – QL-DVSL) sustav koji postavlja ograničenja brzine na svaku prometnu traku zasebno u simulacijskom okruženju u cestovnom mikroskopskom simulacijskom programu SUMO (eng. Simulation of Urban MObility) i programskom okruženju Python. Dobiveni rezultati dokazuju da povećanje udjela autonomnih vozila u prometnom toku djeluje pozitivno, odnosno vožnja se odvija sinkroniziranije, promet se odvija jednolikom brzinom, vrijeme putovanja se smanjuje, postignuta je veća brzina u prometnom toku i smanjena je gustoća vozila čime se smanjuju repovi čekanja.

Abstract (english)

Today's motorways are designed to provide a higher maximum traffic capacity, but the problem of insufficient capacity is increasing. However, urban motorways cannot be widened most of the time, thus, it is necessary to look for other solutions to make better use of the existing capacity of urban motorways. One of the systems from the domain of intelligent transportation systems, which can be used for this, is Variable Speed Limit Control (VSLC). Autonomous vehicles (AV) and connected autonomous vehicles (CAV) have been extensively tested and developed rapidly in the last few years. To examine the characteristics of mixed traffic flows, they must be simulated in different traffic scenarios. In this thesis, a QL-DVSL (Q-Learning – Dual Variable Speed Limit) system that sets separate speed limits on each lane in a simulation environment was created in the road microscopic simulation program SUMO (Simulation of Urban MObility) and the Python programming environment. The obtained results prove that the increase in the share of AV in the traffic flow has a positive effect, i.e., driving is more synchronized, traffic proceeds at a uniform speed, driving time is reduced, and a higher speed has been achieved in the traffic flow, the density of vehicles is reduced, which reduces waiting in queues.

• Keywords: upravljanje promjenjivim ograničenjem brzine; ojačano učenje; autonomna vozila; povezana autonomna vozila; inteligentni transportni sustavi
• Keywords (english): reinforcement learning; variable speed limit control; autonomous vehicle; connected autonomous vehicles; intelligent transport systems
• Language: croatian
• URN:NBN: urn:nbn:hr:119:557628
• Study programme: Title: Intelligent Transport Systems and Logistics; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka prometa (magistar/magistra inženjer/inženjerka prometa)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access
• Created on: 2022-12-09 16:58:40