Analiza repova čekanja i razine uslužnosti urbanih prometnica korištenjem algoritama strojnog učenja i NoSQL baza podataka

master's thesis

• Title (english): Analysis of Queues and Level of Service on Urban Roads Using Machine Learning Algorithms and NoSQL Database
• Author: Leo Tišljarić
• Mentor: Tonči Carić (mentor)
• Committee member: Sadko Mandžuka (predsjednik povjerenstva)
• Committee member: Mario Muštra (član povjerenstva)
• Committee member: Pero Škorput (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: 2018-07-17, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Traffic and Transport Technology; Intelligent Transport Systems and Logistics

Abstract

Traffic data is main starting point for estimating performance or proposing improvements for traffic system. Vehicle speed and control delay are main parameters in Level of service ase-sment for roads and intersections. In this thesis methods for collecting, storing and processing GPS data in order to estimate traffic parameters is proposed. Method for estimating queue length on intersections approaches is presented and machine learning algorithm is used for clustering intersections based on their values of measured performance indicators. For calcula-tions and results, GUI based Python application is developed using MongoDB NoSQL database for storing data and results. Application contains three main features: processing and storing data, calculations of performance measures for intersections and implementation of machine learning algorithm for clustering data.

Abstract (english)

Kvalitetni prometni podaci osnovna su polazišna točka prilikom procjene stanja prometnog sustava ili prijedloga mogućih poboljšanja. Brzina vozila i vrijeme provedeno čekajući na raskrižju osnovni su parametri za procjenu razine uslužnosti neke prometnice ili raskrižja. U ovom radu opisana je metoda prikupljanja i obrade GPS podataka u svrhu procjene osnovnih parametara za opisivanje stanja prometnog toka i razine uslužnosti za prometnice i raskrižja. Prikazana je metoda procjene duljine repova čekanja na semaforiziranim raskrižjima te je primijenjen algoritam strojnog učenja kako bi se grupirala raskrižja na temelju mjerenih mjera učinkovitosti. Za potrebu izračuna svih navedenih parametara izrađena je aplikacija korištenjem Python programskog jezika i MongoDB NoSQL baze podataka. Aplikacija sadrži tri glavne funkcionalnosti: predobrada podataka i unos u bazu podataka, izračun karakterističnih parametara raskrižja i algoritam strojnog učenja za grupiranje podataka.

• Keywords: Machine learning; Queue length; Level of service; clustering; GPS Python; NoSQL; K-means
• Keywords (english): strojno učenje; rep čekanja; razina uslužnosti; grupiranje; GPS; Python; NoSQL; K-means
• Language: croatian
• URN:NBN: urn:nbn:hr:119:716383
• 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: 2018-07-18 12:13:17