Uporaba okrepitvenega učenja za optimizacijo krmiljenja semaforjev

diplomsko delo

Žiga Sušin (Author), Vili Podgorelec (Mentor)

Abstract

V tej nalogi bomo podrobno preučili metodo okrepitvenega učenja in načine implementacije le-tega. Nato ga bomo uporabili za rešitev zadanega problema, ki je optimizacija krmiljenja semaforjev v križišču. V naslednjih poglavjih bomo na splošno opisali strojno učenje, podrobneje pa okrepitveno učenje. Opisali bomo tudi način implementacije v programskem jeziku Python in knjižnice, ki nam pomagajo pri tem. V drugem delu naloge bomo izdelali program s pomočjo pridobljenega znanja. Na koncu pa bomo še predstavili rezultate simulacij.

Keywords

okrepitveno učenje;umetna inteligenca;promet;programski jezik Python;

Data

Language:	Slovenian
Year of publishing:	2021
Typology:	2.11 - Undergraduate Thesis
Organization:	UM FERI - Faculty of Electrical Engineering and Computer Science
Publisher:	[Ž. Sušin]
UDC:	004.85.021:004.43(043.2)
COBISS:	89434371
Views:	263
Downloads:	25
Average score:	0 (0 votes)
Metadata:

Other data

Secondary language:	English
Secondary title:	Reinforcement learning for traffic light control optimization
Secondary abstract:	In this paper we will study the method of reinforcement learning and the ways of its implementation. We will then use the knowledge gained, to solve the given problem, which is optimization of traffic light controls. In chapters that follow, we will describe machine learning in general, and than we will focus more on reinforcement learning and describe it in detail. We will then describe how to implement it in Python programing language and the libraries that help us with its implementation. In the second part of this paper, we will create a program with the acquired knowledge. In the end, we will present the results of the simulation.
Secondary keywords:	Reinforcement learning;artificial intelligence;traffic;Python;
Type (COBISS):	Bachelor thesis/paper
Thesis comment:	Univ. v Mariboru, Fak. za elektrotehniko, računalništvo in informatiko, Informatika in tehnologije komuniciranja
Pages:	VIII, 38 str.
ID:	13333392