Приказ основних података о дисертацији

Improving the efficiency of production processes in the manufacturing industry based on methods of multicriteria analysis and metacheuristics

dc.contributor.advisorTadić, Danijela
dc.contributor.otherAleksić, Aleksandar
dc.contributor.otherMisita, Mirjana
dc.contributor.otherMačužić, Ivan
dc.creatorGojković, Ranka
dc.date.accessioned2022-10-05T20:54:40Z
dc.date.available2022-10-05T20:54:40Z
dc.date.issued2022-01-28
dc.identifier.urihttp://eteze.kg.ac.rs/application/showtheses?thesesId=8555
dc.identifier.urihttps://fedorakg.kg.ac.rs/fedora/get/o:1506/bdef:Content/download
dc.identifier.urihttps://nardus.mpn.gov.rs/handle/123456789/20716
dc.description.abstractProblem procjene i rangiranja grešaka koje mogu dovesti do Lean gubitaka imaju presudan uticaj na efektivnost i pouzdanost proizvodnog kao i ostalih poslovnih procesa preduzeća. U ovoj doktorskoj disertaciji su razvijena dva nova fazi višekriterijumska modela optimizacije zasnovan na Analizi mogućih grešaka i efekata grešaka (prema eng. Failure Mode and Effect Analysis - FMEA), odnosno FMEA okviru za rangiranje grešaka na nivou svakog Lean gubitka. Na samom početku izvršena je identifikacija i grafički prikaz grešaka korišćenjem Išikava dijagrama. Ocjena identifikovanih grešaka vrši se u odnosu na tri faktora rizika (RF) definisana u FMEA metodi. Nedostaci FMEA metode, koje su sugerisani od strane drugih autora, su prevaziđeni kombinacijom ove metode sa pravilima fazi logike i metodama višekriterijumskog odlučivanja (prema eng. Multi Criteria Decision Making - MCDM). U prvom modelu relativna važnost RF i njihove vrijednosti opisani su unaprijed definisanim lingvističkim iskazima koji su modelirani sa trapezoidnim intuitivnim fazi brojevima (prema eng. Trapezoidal intuitionistic fuzzy numbers - TrIFN). Za određivanje vektora težine RF koristi se Analitički hijerarhijski proces proširen sa TrIFN (prema eng. Fuzzy Analytic Hierarchy Process with TrIFN - IFAHP). Rang identifikovanih grešaka daje se upotrebom predložene metode Višekriterijumskog kompromisnog rangiranja proširene sa TrIFN (prema eng. VIKOR with TrIFN - IF-VIKOR). Na kraju, urađena je analiza osjetljivosti koja pokazuje stabilnost predloženog pristupa. U drugom modelu, procjena i rangiranje grešaka koji dovode do Lean gubitaka daju se korišćenjem fazi MCDM metoda proširenih sa intervalnim intuitivnim fazi brojevima (prema eng. Interval valued intuitionistic fuzzy numbers - IVIFN). Relativna važnost RF i njihove vrijednosti opisani su unaprijed definisanim lingvističkim iskazima koji su modelirani sa IVIFN. Modifikovana fazi logika sa pravilima za IVIFN koristi se za određivanje nivoa rizika proizvodnog procesa. U drugom dijelu disertacije, predložen je hibridni model odlučivanja za ocjenu i izbor metoda/tehnika kvaliteta čija primjena dovodi do unaprjeđenja efektivnosti i pouzdanosti proizvodnih procesa u malim i srednjim preduzećima (MSP) prerađivačke industrije. Ovaj model kombinuje FMEA sa trougaoni intuitivni fazi brojevima (prema eng. Triangular intuitionistic fuzzy numbers – TIFN). Sve postojeće neizvjesnosti, relativna važnost RF, njihove vrijednosti, primjenljivost metoda kvaliteta, kao i troškovi primjene opisani su unaprijed definisanim jezičkim iskazima koji su modelirani TIFN. Izbor metoda kvaliteta naveden je kao KP problem, odnosno problem rastegljivog ranca koji se razlaže na potprobleme sa određenim brojem elemenata rješenja. Rješenje ovog problema pronalazi se korišćenjem genetskog algoritma (prema eng. Genetic algorithm - GA) (Gojković et al., 2021). Model je verifikovan kroz studiju slučaja sa podacima iz stvarnog života koji potiču od značajnog broja organizacija iz jednog regiona, čime je pokazan potencijal i primjenljivost razvijenih modela. Pokazano je da su predloženi modeli izuzetno pogodan kao alati za donošenje odluka za poboljšanje efektivnosti i pouzdanosti proizvodnog procesa u MSP prerađivačke industrije.sr
dc.description.abstractThe problem of evaluation and ranking failures that can lead to Lean waste has a critical effect on the safety and reliability of the manufacturing process, and other business processes of enterprises. In this doctoral dissertation, two new fuzzy multicriteria optimization models based on Failure Mode and Effect Analysis - FMEA have been developed to rank failures at the level of each Lean waste. At the beginning, failures were identified using the Ishikawa diagram. The evaluation of the identified failures is performed in relation to the three risk factors (RF) defined in the FMEA method. The disadvantages of the FMEA method, which have been suggested by other authors, have been overcome by combining this method with the fuzzy logic rols and the Multi Criteria Decision Making (MCDM). In the first model, the relative importance of RF and their values are described by predefined linguistic statements modeled with trapezoidal intuitionistic fuzzy numbers (TrIFN). The Fuzzy Analytic Hierarchy Process with TrIFN (IF-AHP) was used to determine the RF weight vector. The rank of identified failures is given using the proposed VIKOR with TrIFN (IF-VIKOR). Finally, a sensitivity analysis was performed showing the stability of the proposed approach. In the second model, estimation and ranking of failures leading to Lean waste are given using the fuzzy MCDM with interval valued intuitionistic fuzzy numbers (IVIFN). The relative importance of RF and their values are described by predefined linguistic statements modeled with IVIFN. A modified fuzzy logic ruls with IVIFN rules is used to determine the level of risk of the production process. In the second part of the dissertation, a hybrid decision-making model for evaluation and selection of quality methods/techniques is proposed, the application of which leads to the improvement of efficiency and reliability of production processes in small and medium enterprises (SMEs) of the manufacturing industry. This model combines FMEA with the triangular intuitionistic fuzzy numbers (TIFN). All existing uncertainties, the relative importance of RF, their values, the applicability of quality methods/techniques, as well as the costs of application are described by pre-defined linguistic statements modeled by TIFN. The choice of quality methods/techniques is stated as a KP problem. It is a Rubber Knapsack problem that decomposes into subproblems with a certain number of solution elements. The solution to this problem is found using a genetic algorithm (GA) (Gojković et al., 2021). The model was verified through a case study with real life data originating from a significant number of organizations from one region, showing the potential and applicability of the developed models. He showed that the proposed models are extremely suitable as decision-making tools for improving the efficiency and reliability of the production process in the SME manufacturing industry.en
dc.formatapplication/pdf
dc.languagesr
dc.publisherУниверзитет у Крагујевцу, Факултет инжењерских наукаsr
dc.rightsopenAccessen
dc.sourceУниверзитет у Крагујевцуsr
dc.subjectLean gubicisr
dc.subjectLean wasteen
dc.subjectgreškesr
dc.subjectintuitivni fazi skupovisr
dc.subjectFMEAsr
dc.subjectfazi MCDMsr
dc.subjectIF-AHPsr
dc.subjectIF – VIKORsr
dc.subjectIF – TOPSISsr
dc.subjectizbor metoda/tehnika kvalitetasr
dc.subjectGAsr
dc.subjectfailuresen
dc.subjectintuitionistic fuzzy setsen
dc.subjectFMEAen
dc.subjectfuzzy MCDMen
dc.subjectIFAHPen
dc.subjectIF – VIKORen
dc.subjectIF – TOPSISen
dc.subjectselection of quality methodsen
dc.subjectGAen
dc.titleУнапрјеђење ефективности процеса производње у прерађивачкој индустрији засновано на методама вишекритеријумске анализе и метахеуристикеsr
dc.title.alternativeImproving the efficiency of production processes in the manufacturing industry based on methods of multicriteria analysis and metacheuristicsen
dc.typedoctoralThesis
dc.identifier.fulltexthttp://nardus.mpn.gov.rs/bitstream/id/146186/Ranka_Gojkovic_FIN.pdf
dc.identifier.fulltexthttp://nardus.mpn.gov.rs/bitstream/id/146187/Doctoral_thesis_12614.pdf
dc.identifier.rcubhttps://hdl.handle.net/21.15107/rcub_nardus_20716


Документи за докторску дисертацију

Thumbnail
Thumbnail

Ова дисертација се појављује у следећим колекцијама

Приказ основних података о дисертацији