Okruženje za simulacije sustava za upravljanje promjenjivih ograničenja brzine na gradskim autocestama zasnovano na dubokom učenju

master's thesis

• Title (english): Framework for Simulation of Variable Speed Limit Control Systems on Urban Motorways Based on Deep Learning
• Author: Filip Vrbanić
• Mentor: Edouard Ivanjko (mentor)
• Committee member: Sadko Mandžuka (predsjednik povjerenstva)
• Committee member: Martin Gregurić (član povjerenstva)
• Committee member: Miroslav Vujić (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: 2019-09-25, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Traffic and Transport Technology; Intelligent Transport Systems and Logistics

Abstract

Radi povećane prometne potražnje dolazi do nastanka zagušenja na gradskim autocestama te se javlja potreba za uvođenjem upravljanja prometom. Upravljanje prometom ima za cilj smanjenje i prevenciju posljedica nastajanja prometnih zagušenja. Sustav za upravljanje ograničenjem brzine na gradskim autocestama predstavlja jedno od mogućih rješenja iz područja inteligentnih transportnih sustava. Takav pristup u upravljanju promjenjivim ograničenjem brzine na gradskim autocestama može koristiti algoritme zasnovane na dubokom učenju. U ovom radu je opisano upravljanje prometom na gradskim autocestama zasnovano na upravljanju promjenjivim ograničenjem brzine, analizirana je mogućnost primjene dubokog učenja s posebnim naglaskom na algoritmu učenja dualne duboke Q-mreže u upravljanju promjenjivim ograničenjem brzine te je izrađeno okruženje za simulaciju navedenog upravljanja korištenjem mikroskopskog simulatora VISSIM i programskog okruženja Python/Keras. Dobiveni rezultati dokazuju da je primjenom dubokog učenja moguće iskoristiti informaciju o brzini i položaju svakog vozila za povećanje razine uslužnosti gradske autoceste.

Abstract (english)

In the case of congestion on urban motorways due to increased traffic demand, there is a need to introduce dynamic traffic management to reduce and prevent the effects of traffic congestion. The urban motorway speed limit control system is one of the possible solutions in the field of Intelligent Transport System. Such an approach in managing variable speed limits on urban motorways can utilize a deep learning algorithm. This thesis describes urban motorway traffic management based on the variable speed limit control, analyzes the possibility of applying deep learning algorithm with special emphasis on the dueling deep Q-network learning algorithm in variable speed limit control, and creating an environment for simulating such system using the microscopic simulator VISSIM and the Python/Keras programming environment. The results prove that by applying such an algorithm it is possible to use information about the speed and position of each vehicle to increase the level of service of the urban motorway.

• Keywords: upravljanje promjenjivim ograničenjem brzine; simulacijsko okruženje; duboko učenje; ojačano učenje; inteligentni transportni sustavi
• Keywords (english): variable speed limit control; framework for simulation; deep learning; reinforcement learning; intelligent transport systems
• Language: croatian
• URN:NBN: urn:nbn:hr:119:060236
• 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: 2019-11-15 12:29:13