Cooperative Ramp Metering for Urban Motorways Based on Machine Learning

doctoral thesis

• Title (croatian): Kooperativno upravljanje priljevnim tokovima na urbanim autocestama zasnovano na strojnom učenju
• Author: Martin Gregurić
• Mentor: Sadko Mandžuka (mentor)
• Mentor: Edouard Ivanjko (komentor)
• Committee member: Davor Brčić (predsjednik povjerenstva)
• Committee member: Sadko Mandžuka (član povjerenstva)
• Committee member: Edouard Ivanjko (član povjerenstva)
• Committee member: Niko Jelušić (član povjerenstva)
• Committee member: Ivan Petrović (član povjerenstva)
• Granter: University of Zagreb; Faculty of Transport and Traffic Sciences; Zagreb
• Defense date and country: 2018-04-24, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Traffic and Transport Technology; Road and Rail Transport
• Universal decimal classification (UDC): 656.1/.5 - Land transport

Abstract

To cope with today’s urban motorway congestions and the inability to increase motorway capacity in urban environments requires the implementation of advanced control methods. These methods are an integral part of Intelligent Transportation Systems (ITS). An ITS essentially integrates information and communication technology to solve the congestion problems. Ramp metering (RM) and Variable Speed Limit Control (VSLC) are some of the most widely used urban motorway traffic control methods. RM... More provide direct influence over the on-ramp flows by using specialized traffic lights, while the VSLC control speed of mainstream flow by using variable messaging signs. A dedicated algorithm for RM or VSLC uses sensory data form an urban motorway to compute actions that will have a positive impact on both types of traffic flow. This study will focus on the cooperation of an RM and a VSLC systems, and the integration of several different RM algorithms into a single algorithm called INTEGRA. The algorithm is created by using the Adaptive Neuro-fuzzy Inference System (ANFIS) as an instance of machine learning techniques. Furthermore, INTGERA is expanded in order to integrate its original functionality with a recurrent neural network for traffic demand prediction. As the final step, this doctoral thesis will provide evaluation of different criteria for learning dataset functional setup, based on which ANFIS neural network of INTEGRA will be learned. Results of all mentioned approaches will be compared and discussed in relation with other commonly used urban motorway control methods. Less

Abstract (croatian)

Glavnina istraživanja u ovom doktorskom radu vezana je upravo za upravljanje priljevnim tokovima s posebnim naglaskom na kooperaciju s drugim sustavima upravljanja prometom, te primjeni strojnog učenja. Također, u kooperaciji s upravljanjem priljevnih tokova razmatrat će se druge upravljačke metode kao što su sustav zabrane prometovanja određenim prometnim trakama, te potpuno ili djelomično upravljanje vozilima opremljenim posebnim računalnim jedinicama. Od strane autora predložen je... More neuro-neizraziti okvir za učenje koji omogućuje integraciju različitih strategija upravljanja priljevnim tokovima. CTMSIM makro-simulacijski alat koji je izrađen u Matlab programskom okruženju korišten je u simulaciji odabranih metoda upravljanja prometom na urbanim autocestama. Simulator je proširen od strana autora kako bi podržao kooperativno upravljanje priljevnim tokovima, kao i sustav za promjenjivo ograničenje brzina vozila. Less

• Keywords: cooperative control; ramp metering; variable speed limit control; urban motorways; adaptive neuro fuzzy inference systems; recurrent neural network; learning dataset
• Keywords (croatian): Rad ne sadrži ključne riječi na drugom jeziku.
• Language: english
• URN:NBN: urn:nbn:hr:119:512923
• Study programme: Title: Technological systems in traffic and transport; Study programme type: university; Study level: postgraduate; Academic / professional title: doktor/doktorica znanosti, područje tehničkih znanosti, polje tehnologija prometa i transport (doktor/doktorica znanosti, područje tehničkih znanosti, polje tehnologija prometa i transport)
• Type of resource: Text
• Extent: XVI, 144 str. ; 30 cm
• File origin: Born digital
• Access conditions: Open access
• Created on: 2019-03-27 12:12:20