Optimiranje prometnega toka semaforiziranega križišča

magistrsko delo

Šemso Hrnjičić (Author), Uroš Rajkovič (Mentor)

Abstract

V človeški naravi je, da izboljšujemo vse procese, ki jih lahko. Ko vidimo, da nekateri sistemi ali procesi ne dosežejo svojega potenciala, sanjarimo, kako bi to lahko spremenili in izboljšali, četudi je to le podzavestno. Promet družba ves čas spreminja, nadgrajuje in izboljšuje v dobro vseh. Dolgo časa sem razmišljal o dinamičnem centraliziranem sistemu, ki bi spremljal stanje na cestah v določeni regiji. Prilagajal bi se spremembam na cestah v realnem času in bi imel določene prioritete. Iz tega razmišljanja je nastala ideja za magistrsko delo. Cilj magistrskega dela je eksperimentirati in ustvariti simulacijo semaforiziranega križišča s spremembami, ki smo si jih zamislili, da bi ugotovili, ali so smiselne. V našem primeru je to eno samo križišče. Križišče smo izbrali in ga rekreirali v 3D okolju Blender. Potem ko smo križišče vizualizirali, smo se lotili razvoja simulacije v Unity. Sprogramirali smo logiko agentov, ki bodo v vlogi vsakdanjih voznikov, odločanje agentov pri semaforiziranem križišču in delovanje semaforiziranega križišča. Za simulacijo smo se omejili na določeno semaforizirano križišče in na promet ob določenem času. Preden smo lahko simulirali promet, smo morali prešteti vsa vozila iz vseh smeri. Ko smo dobili vse podatke in ko smo imeli pripravljeno simulacijo, smo lahko simulirali promet, optimizirali križišče na različne načine in rezultate simulacij primerjali. Z magistrskim delom je prikazan razvoj simulacijskega okolja, vizualizacija trenutnega stanja in primerjava različnih načinov izboljševanja. Na koncu smo začeli diskusijo o umetni inteligenci in strojnem učenju ter o pomenu takih tehnologij za razvoj in izboljšavo procesov in sistemov, ki jih uporabljamo vsak dan.

Keywords

vizualizacija;razvoj;simulacija;optimizacija;strojno učenje;

Data

Language:	Slovenian
Year of publishing:	2022
Typology:	2.09 - Master's Thesis
Organization:	UM FOV - Faculty of Organizational Sciences
Publisher:	[Š. Hrnjičić]
UDC:	625
COBISS:	114247171
Views:	82
Downloads:	8
Average score:	0 (0 votes)
Metadata:

Other data

Secondary language:	English
Secondary title:	Optimizing traffic flow in an intersection with traffic lights
Secondary abstract:	It is in the human nature to improve all the processes we can. When we see that a certain system or process is not reaching their full potential, we dream of how we could change and improve it. Even if it is only subconscious. Traffic is one of the things that our society is constantly changing, upgrading and improving for the benefit of all. For a long time, I dreamed of a centralized system that would monitor the situation on the roads in an region. A system that would be dynamic, that would adapt to real-time changes, and that would have priorities. From this daydreaming, an idea for a master’s thesis arose. The goal of the master’s thesis is to experiment and start at the lowest level and see if the changes, that we would implement at a traffic light, would make sense at all. In our case, this is a single intersection. We chose a intersection and recreated it in an 3D environment in Blender. After we finished the visualisation, we started developing the simulation in Unity. We programmed the logic of the agents who take on the role of everyday drivers, the decision-making of the agents at the traffic lights and the traffic light logic. For the simulation, we limited ourselves to a certain traffic light intersection and to traffic at a certain time. Before we could simulate traffic, we had to count all the vehicles at that intersection from all directions. Once we got all the data and we had the simulation ready, we were then able to simulate the traffic, optimize the intersection in different ways and compare the results. The master’s thesis presents the development of the simulation environment, visualization of the current situation and the comparison between the different ways of improvements. Finally, we had a discussion on artificial intelligence, machine learning and its improvement of processes and systems that we use every day.
Secondary keywords:	Prometna signalizacija;Univerzitetna in visokošolska dela;
Type (COBISS):	Master's thesis/paper
Thesis comment:	Univ. v Mariboru, Fak. za organizacijske vede
Pages:	VI, 53 f.
ID:	15638016