delo diplomskega seminarja
Eva Erzin (Author), Alen Orbanić (Mentor), Bogdan Filipič (Co-mentor)

Abstract

Večkriterijski optimizacijski problemi so del vsakdana. Včasih jih uspemo rešiti sami, včasih pa so prezahtevni in za to potrebujemo pomoč. Dober pristop k reševanju večkriterijskih optimizacijskih problemov so genetski algoritmi. V tem delu se ukvarjamo z večkriterijskimi optimizacijskimi problemi z omejitvami. Najprej jih definiramo in opišemo njihovo rešitev - Pareto optimalno množico. Nato predstavimo genetske algoritme, si podrobneje ogledamo dva izmed njih, NSGA-II in MOEA/D ter pregledamo obstoječe načine obravnavanja omejitev v večkriterijski optimizaciji, s katerimi lahko genetske algoritme za večkriterijsko optmizacijo prilagodimo tako, da lahko obravnavajo tudi probleme z omejitvami. Na koncu predstavimo še dva testna večkriterijska optimizacijska problema z omejitvami, na njima preizkusimo prej predstavljena algoritma ter dva izmed načinov obravnavanja omejitev in rezultate interpretiramo.

Keywords

matematika;večkriterijska optimizacija z omejitvami;genetski algoritmi;NSGA-II;MOEA/D;

Data

Language: Slovenian
Year of publishing:
Typology: 2.11 - Undergraduate Thesis
Organization: UL FMF - Faculty of Mathematics and Physics
Publisher: [E. Erzin]
UDC: 519.8
COBISS: 18437465 Link will open in a new window
Views: 991
Downloads: 254
Average score: 0 (0 votes)
Metadata: JSON JSON-RDF JSON-LD TURTLE N-TRIPLES XML RDFA MICRODATA DC-XML DC-RDF RDF

Other data

Secondary language: English
Secondary title: Constraint handling in multiobjective optimization
Secondary abstract: Multiobjective optimization problems are a part of everyday life. Sometimes we manage to solve them and other times they prove to be too difficult and we need help solving them. A good approach to solving multiobjective optimization problems are genetic algorithms. In this work we deal with constrained multiobjective problems. First we describe them and their solution - the Pareto front. Then we present genetic algorithms, desribe two of them, NSGA-II and MOEA/D, more in-depth and review existing constraint handling methods, that allow us to adapt existing multiobjective genetic algorithms for constrained multiobjective optimization. Finally we present two multiobjective constrained test problems, use them to test the beforementioned genetic algorithms and two of the constraint handling techniques, and interpret the results.
Secondary keywords: mathematics;constrained multiobjective optimization;genetic algorithms;NSGA-II;MOEA/D;
Type (COBISS): Final seminar paper
Study programme: 0
Thesis comment: Univ. v Ljubljani, Fak. za matematiko in fiziko, Oddelek za matematiko, Matematika - 1. stopnja
Pages: 25 str.
ID: 10959927