Primerjava različnih algoritmov po vzoru obnašanja netopirjev

diplomsko delo

Žan Grajfoner (Author), Iztok Fister (Mentor), Lucija Brezočnik (Co-mentor)

Abstract

V diplomski nalogi smo se osredotočili na algoritme po vzorih iz narave. Opisujemo evolucijske algoritme, katerih navdih je bila Darwinova teorija o boju za obstanek, in algoritme inteligence roja, ki črpajo navdih iz obnašanja roja živali v naravi. Primerjali smo osnovni algoritem po vzoru obnašanja netopirjev in hibridno različico algoritma po vzoru obnašanja netopirjev. Raziskali smo razlike med osnovnima arhitekturama obeh algoritmov, pripadajoče parametre, kot tudi področja uporabe obeh algoritmov. Primerjavo smo izvedli na praktičnem primeru optimizacije desetih testnih funkcij na treh različnih dimenzijah problema (10, 20, 30). Prav tako smo raziskali vpliv različnih velikosti populacije (20, 30, 50) pri obeh algoritmih. Ugotovili smo, da so rezultati optimizacije hibridne različice algoritma boljši od standardne različice algoritma.

Keywords

algoritem po vzoru obnašanja netopirjev;evolucijski algoritmi;hibridizacija;inteligenca roja;računska inteligenca.;diplomske naloge;

Data

Language:	Slovenian
Year of publishing:	2019
Typology:	2.11 - Undergraduate Thesis
Organization:	UM FERI - Faculty of Electrical Engineering and Computer Science
Publisher:	[Ž. Grajfoner]
UDC:	004.421(043.2)
COBISS:	22577174
Views:	1043
Downloads:	142
Average score:	0 (0 votes)
Metadata:

Other data

Secondary language:	English
Secondary title:	A comparison between different bat algorithms
Secondary abstract:	In this thesis, we are focusing on nature-inspired algorithms. We describe evolutionary algorithms inspired by the Darwinian theory and swarm intelligence algorithms that have been inspired by the behaviour of swarms in nature. We compare the original bat algorithm with the hybrid bat algorithm and investigate the differences between the regular architecture of both algorithms, related parameters, and areas of use of both algorithms. In the experiment, we use ten benchmark functions on three different dimensions (10, 20, 30). We also research the influence of the population size (20, 30, 50) on both algorithms. Results show that the hybrid bat algorithm outperforms the standard bat algorithm.
Secondary keywords:	bat algorithm;evolutionary algorithms;hybridization;swarm intelligence;computational intelligence.;
Type (COBISS):	Bachelor thesis/paper
Thesis comment:	Univ. v Mariboru, Fak. za elektrotehniko, računalništvo in informatiko, Informatika in tehnologije komuniciranja
Pages:	VIII, 44 f.
ID:	11204317