izpis_h1_title_alt

Obdelava kompleksnih dogodkov pri spremljanju proizvodnega procesa
KRIVEC, TADEJ (Author), Mušič, Gašper (Mentor) More about this mentor... This link opens in a new window, Gradišar, Dejan (Co-mentor)

.pdfPDF - Presentation file, Download (2,92 MB)

Abstract
V magistrski nalogi se osredotočam na procesiranje velikih tokov podatkov v realnem času. Zaradi zahtev po optimizaciji celotnega poslovnega procesa je treba pridobiti informacije v realnem času s čim manjšo časovno zakasnitvijo. Tako se zmanjšajo stroški poslovanja in vodenja proizvodnje. Na proizvodnem nivoju se obdelava kompleksnih dogodkov (ang. complex event processing) uporablja za zaznavanje napak in zagotavljanja kakovosti. Sistemi za statistično spremljanje veličin lahko zaznavajo nenormalne pogoje in pošiljajo alarme. Obdelava kompleksnih dogodkov deluje tako, da lahko sprejme različne tokove podatkov, od poslovnih do proizvodnih in omogoča preproste poizvedbe na širokem naboru podatkov. Poizvedbe so statične, podatki pa dinamični, kar pomeni, da surovih podatkov ne shranjujemo neposredno v bazo podatkov. Tak način omogoča najmanjšo časovno zakasnitev podatkov za odločanje, procesirane podatke pa lahko še vedno shranimo za kasnejšo obdelavo v obliki podatkovnega jezera (ang. data lake). Rešitev je implementirana z platformo Microsoft StreamInsight. Podatki so zgodovinsko in realnočasno analizirani s sistemom za upravljanje podatkovnih baz (ang. database management system) Microsoft SQL Server ter orodjem za poslovno analitiko (ang. business intelligence) Power BI. Predstavljena je rešitev za zaznavanje napak z metodo PCA (ang. principal component analysis) in DPCA (ang. dynamic principal component analysis). Nesprotni del algoritma je izračunan v programskem jeziku Python, kjer se uporablja Jupyter Notebook. Sprotni del je realiziran na platformi Microsoft StreamInsight z uporabo knjižnice MathNet jezika C#.

Language:Slovenian
Keywords:obdelava kompleksnih dogodkov, zaznavanje napak, PCA, DPCA, tok podatkov, podatkovno jezero, poslovna analitika, Microsoft StreamInsight, Microsoft SQL Server, Power BI, Python
Work type:Master's thesis/paper (mb22)
Organization:FE - Faculty of Electrical Engineering
Year:2018
Views:275
Downloads:156
Metadata:XML RDF-CHPDL DC-XML DC-RDF
 
Average score:(0 votes)
Your score:Voting is allowed only to logged in users.
:
Share:AddThis
AddThis uses cookies that require your consent. Edit consent...

Secondary language

Language:English
Title:Complex event processing in production process monitoring
Abstract:
This master thesis focuses on the processing of big data streams with high throughput in real time. Because of the demands for optimization of the business process, data has to be obtained in real time or with minimal time delay. On the production level, complex event processing is mostly used for fault detection and quality assurance. Statistical process control systems can also notify out-of-bound conditions and send alarms. Complex event processing allows for querying different types of data streams, from business to production, with relative ease. Queries are static and data is dynamic, which is a paradigm shift from the conventional analysis of static datasets. It enables real time querying without the delay of writing the data in the database. This kind of analysis guarantees minimal time delay for decision making and processed data can still be stored in the form of a data lake. The complex event processing solution is implemented on the platform Microsoft StreamInsight. Real time and historical data is analysed in the database management system Microsoft SQL Server and Microsoft platform for business intelligence solutions Power BI. Solution is proposed for fault detection with the use of PCA (principal component analysis) and DPCA (dynamic principal component analysis). Offline part of the algorithm is implemented in Python with the use of Jupyter Notebook. Online part is implemented on the platform StreamInsight with the MathNet library of the programming language C#.

Keywords:complex event processing, fault detection, PCA, DPCA, data streams, data lake, business intelligence, Microsoft StreamInsight, Microsoft SQL Server, Power BI, Python

Similar documents

Similar works from RUL:
Similar works from other Slovenian collections:

Comments

Leave comment

You have to log in to leave a comment.

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

Back