Stvarnovremenska analiza podataka korisnika javnog gradskog prijevoza

master's thesis

• Title (english): Real-time data analysis of public transport users
• Author: Petar Lazić
• Mentor: Mario Kušek (mentor)
• Committee member: Mario Kušek (predsjednik povjerenstva)
• Committee member: Gordan Ježić (član povjerenstva)
• Committee member: Krunoslav Tržec (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2021-07-08, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Tema rada je rješenje za problem detektiranja stanja prometa u gradskoj okolini, prvenstveno stvarnovremenskog lociranja vozila javnog prijevoza. Zamišljeno rješenje uključuje korisnike dijele svoju lokaciju te opcionalno i druge podatke poput identifikatora vozila u kojem se nalaze. Zatim sustav na temelju tih podataka određuje lokacije vozila što daljnjom analizom omogućuje određivanje vremena dolazaka na pojedine stanice. Za razliku od drugih rješenja koja bi mogla zahtijevati znatna ulaganja u gradsku infrastrukturu, jedino što ovakav sustav treba je mobilna aplikacija, poslužitelj te dovoljan broj pojedinaca koji su voljni sudjelovati. Radi se o pojmu grupnog opažanja okoline (engl. crowdsensing). U radu su bili obrađeni problemi poput grupiranja korisnika, detekcija o kojoj se ruti radi, izbacivanje uljeza (engl. outliera) iz grupe, itd.

Za potrebe rada je bila razvijena simulacija koja se temeljila na podacima dobivenim iz GTFS datoteka. Analizom tih datoteka se došlo do informacija kao što su vremena putovanja između parova stanica za svaki tramvaj, udaljenost između stanica, popisi tramvaja koji prometuju određenom stanicom.

Osim kretanja tramvaja, simulacija također uključuje simuliranje korisnika koji se voze tim tramvajima. Ukratko, simulacija radi tako da sekvencijalno upravlja redom tramvajima čije se kretanje simulira linearnom interpolacijom, stanicama koje zaprimaju dolaske i odlaske tramvaja te dojavljuju korisnicima i korisnicima koji se nasumično stvaraju na stanicama te čekaju određene tramvaje. Sustav za predviđanje lokacija tramvaja od korisnika dobiva samo lokaciju te opcionalno broj tramvaja i smjer te na temelju tih dobivenih podataka grupira korisnike u grupe koje predstavljaju tramvaje i povremeno provjerava ispravnost grupa.

Abstract (english)

The topic of this paper is a solution to the problem of detecting traffic state in the urban environment, primarily the real-time location of public transport vehicles.

The imagined solution involves users sharing their location and optionally other data such as the identifier of the vehicle in which they are currently driving in. Then, based on this data, the system determines the locations of vehicles, which allows further analysis to determine the time of arrival at individual stations. Unlike other solutions that could require significant investment in the city infrastructure, the only thing such a system needs is a mobile application, a server and a sufficient number of individuals willing to participate. The name of this technique is called crowdsensing. The paper deals with problems such as user grouping, detection of the route based on location history, detection of outliers in a group, etc.

For the purposes of this paper, a simulation based on data obtained from GTFS files was developed. The analysis of these files yielded information such as travel times between pairs of stations for each tram, the distance between stations and lists of trams running on a particular station.

In addition to the movement of trams, the simulation also includes the simulation of users riding these trams. In short, the simulation works by sequentially controlling trams which move by a method linear interpolation, stations that get notified by tram arrivals/departures and finally users who get randomly created at stations and wait for certain trams. The tram location prediction system receives only the location and optionally the tram identifier from the user, and on based on this data it groups users into groups representing trams and periodically checks the correctness of the groups.

• Keywords: Internet stvari; Usluge suradnog opažanja okoline; Javni prijevoz
• Keywords (english): Internet of Things; Crowdsensing; Public transport
• Language: croatian
• URN:NBN: urn:nbn:hr:168:485173
• Study programme: Title: Computing; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-01-13 12:52:33