| | SLO | ENG | Cookies and privacy

Bigger font | Smaller font
First page > Show document

Show document Help

Title:Razporejanje proizvodnje z metodami strojnega učenja
Authors:ID Rupnik, Robert (Author)
ID Kofjač, Davorin (Mentor) More about this mentor... New window
ID Brezavšček, Alenka (Comentor)
Files:.pdf MAG_Rupnik_Robert_2016.pdf (3,25 MB)
MD5: A6946C9727AD0AFF649D5E7C15173AA6
 
Language:Slovenian
Work type:Master's thesis/paper
Organization:FOV - Faculty of Organizational Sciences in Kranj
Abstract:Procesi v kompleksnih proizvodnih okoljih postajajo vse bolj nepredvidljivi in se zaradi nenehnih spreminjajočih se zahtev odjemalcev v globalnem okolju čedalje hitreje spreminjajo. Podjetja so se tako že v preteklosti pričela podrobneje organizirati in so v procese vključevala pomagala za terminiranje proizvodnje posameznega izdelka. Danes so na trgu že zelo dodelani računalniško podprti programi, ki pa žal ne predstavljajo ustrezne rešitve v kompleksnih okoljih, kjer imamo opravka z masovnimi, stohastičnimi tokovi materialov. V nalogi smo prikazali praktično uporabo aplikacije, izdelane z metodo strojnega učenja in genetskih algoritmov, v konkretnem proizvodnem okolju jeklarne SIJ Acroni d. o. o. Podjetje sestavljajo štiri enote, optimirali pa smo sklop strojev v eni izmed njih. Zaradi kompleksnosti proizvodnje izključno unikatnih izdelkov proces optimiranja v takem primeru preseže orodja klasičnega terminiranja kakor tudi človeško kombinatoriko. Reševanja izziva zmanjšanja zastojev smo se lotili z uporabo znanja s področja umetne inteligence in genetskih algoritmov. Razvili smo model za sklop strojev in izvedli njegovo validacijo s pomočjo dogodkovne simulacije. Genetske algoritme smo uporabili za iskanje optimalnega proizvodnega razporeda. V izvedeni preliminarni študiji smo ugotovili, da lahko z uporabo genetskih algoritmov čas proizvodnje skrajšamo v povprečju tudi za 4 %, kar pomeni velike časovne prihranke in za podjetje tudi nižje stroške obratovanja proizvodnje. Na ta način smo dokazali, da predstavljajo genetski algoritmi primerno metodo za optimiranje kompleksnih proizvodnih procesov, kar pripomore k večji produktivnosti proizvodnega procesa.
Keywords:strojno učenje, genetski algoritmi, optimiranje, stohastični procesi, planiranje (terminiranje) proizvodnega procesa, proizvodni management
Place of publishing:Maribor
Year of publishing:2016
PID:20.500.12556/DKUM-63462 New window
COBISS.SI-ID:7794963 New window
NUK URN:URN:SI:UM:DK:LL21PNHH
Publication date in DKUM:17.10.2016
Views:2227
Downloads:153
Metadata:XML DC-XML DC-RDF
Categories:FOV
:
RUPNIK, Robert, 2016, Razporejanje proizvodnje z metodami strojnega učenja [online]. Master’s thesis. Maribor. [Accessed 20 March 2025]. Retrieved from: https://dk.um.si/IzpisGradiva.php?lang=eng&id=63462
Copy citation
  
Average score:
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
(0 votes)
Your score:Voting is allowed only for logged in users.
Share:Bookmark and Share


Hover the mouse pointer over a document title to show the abstract or click on the title to get all document metadata.

Secondary language

Language:English
Title:Production scheduling with machine learning methods
Abstract:The processes in complex production environments are becoming more and more unpredictable and are changing much more quickly due to the constantly changing requirements of the customers on a global scale. This is the reason why already years ago, companies started to organise their work processes in much more detail, for example with the use of software for production scheduling in their production processes. Today, various sophisticated computerised programmes are available on the market; however, they do not provide effective solutions for complex environments with massive stochastic material flows. The following thesis presents the practical use of an application created on the basis of a machine learning method and a method of genetic learning within the production environment of the company SIJ Acroni, Ltd. The company comprises of four units, one of which was subject to our optimization process. The production in this unit is extremely complex due to the fact that most products are unique. This is also the reason why the optimisation process exceeds the use of typical production scheduling methods. We addressed the challenge of how to decrease the stoppage with the use of knowledge from the field of artificial intelligence and genetic algorithms. A model for a specific machine group was developed and validated with event simulation method. We used genetic algorithms for the determination of the optimum production plan. The result of the survey shows that, with the use of genetic algorithm method, production time can on average be reduced by four per cent, resulting in a substantial reduction of time needed for the production process and consequently also lower production costs. We thus proved that genetic algorithm methods could be used effectively not only for optimising complex production processes, but also for increasing the productivity of the production process.
Keywords:machine learning, genetic algorithm, process optimization, production scheduling, stochastic process, production management


Comments

Leave comment

You must log in to leave a comment.

Comments (0)
0 - 0 / 0
 
There are no comments!

Back
Logos of partners University of Maribor University of Ljubljana University of Primorska University of Nova Gorica