doctoral thesis
Clustering of objects with spatial uncertainty applied on a warehouse model

Mirko Kohler (2014)
Sveučilište Josipa Jurja Strossmayera u Osijeku
Fakultet elektrotehnike, računarstva i informacijskih tehnologija Osijek
Zavod za programsko inženjerstvo
Katedra za vizualno računarstvo
Metadata
TitleRazvrstavanje objekata s prostornom neodređenosti primijenjeno na modelu skladišta
AuthorMirko Kohler
Mentor(s)Krešimir Nenadić
Abstract
U ovoj disertaciji predstavljene su prilagođene metode i predložena je nova Kombinirana metoda za razvrstavanje objekata koji sadrže neodređenost u svojim prostornim podacima u ℝ3 geometrijskom prostoru. Kombinirana metoda koristi prednosti i uklanja nedostatke postojećih metoda. Ona omogućuje brže razvrstavanje objekata neodređenosti koristeći jednostavne matematičke funkcije usporedbe za odbacivanje grozdova, a koristeći simetralne ravnine za usporedbu i odbacivanju grozdova kao potencijalnih kandidata odbacuje dodatni broj grozdova. Metoda znatno ubrzava proces odbacivanja grozdova, u odnosu na postojeće metode, tako što minimalizira ukupan broj objekata za koji je potrebno računati očekivanu udaljenost. Stvoren je model skladišta za verifikaciju nove metode razvrstavanja i u njemu provedeni pokusi s ciljem skraćivanja vremena razvrstavanja proizvoda i pronalaženja optimalnih pozicija za poslužitelje. Predstavljeni su prijedlozi za optimiranje posluživanja prilagođavanjem modela skladišta novoj metodi. Na temelju prijedloga kreiran je novi model skladišta nad kojim su provedeni isti pokusi s ciljem dokazivanja ubrzavanja procesa razvrstavanja.
Keywordscluster pruning clustering uncertain objects optimization warehouse
Parallel title (English)Clustering of objects with spatial uncertainty applied on a warehouse model
Committee MembersŽeljko Hocenski (committee chairperson)
Mario Žagar (committee member)
Alfonzo Baumgartner (committee member)
Damir Blažević (committee member)
GranterSveučilište Josipa Jurja Strossmayera u Osijeku
Fakultet elektrotehnike, računarstva i informacijskih tehnologija Osijek
Lower level organizational unitsZavod za programsko inženjerstvo
Katedra za vizualno računarstvo
PlaceOsijek
StateCroatia
Scientific field, discipline, subdisciplineTECHNICAL SCIENCES
Computing
Program Engineering
UDK62
APPLIED SCIENCES. MEDICINE. TECHNOLOGY
Engineering. Technology in general
Study programme typeuniversity
Study levelpostgraduate
Study programmePostgraduate doctoral study programme; branch Power Engineering and Communications, branch Informatics
Study specializationbranch Informatics
Academic title abbreviationdr.sc.
Genredoctoral thesis
Language Croatian
Defense date2014-11-13
Promoted date2016-04-09
Parallel abstract (English)
In this dissertation adapted methods and new combined method for clustering spatially uncertain data in ℝ3 geometrical space are presented. The proposed combined method uses the advantages and eliminates the disadvantages of the existing methods. Combined method enables faster clustering of objects with spatial uncertainty using simple comparisons mathematical functions for cluster pruning. Pruning additional number of clusters is achived by using bisector planes for comparison and rejection of clusters as potential candidates. The new method significantly speeds up the process of pruning clusters, compared to existing methods. This is achived by minimizing the total number of objects for which it is necessary to calculate the expected distance. A warehouse model for verification of new clusternig method is created. Experiments are carried out with the aim of shortening the time for clustering of products and finding the optimal position for automated servers. Proposals for servis optimizing by customizing warehouse model, for new method, are presented. On the basis of proposals new warehouse model is created. On the new developed model same experiments are performed with the aim of proving acceleration of the clustering process.
Parallel keywords (Croatian)objekti s neodređenosti optimizacija odbacivanje grozdova razvrstavanje skladište
Extent108
Versionaccepted version
Resource typetext
Access conditionOpen access
Terms of usehttp://rightsstatements.org/vocab/InC/1.0/
Noteaccepted version
URN:NBNhttps://urn.nsk.hr/urn:nbn:hr:200:277198
CommitterAnka Ovničević