Izradba strukturnih karata pomoću neuronskih mreža iz bušotinskih i seizmičkih podataka - primjer u naftno-plinskom polju Kloštar

master's thesis

• Title: Izradba strukturnih karata pomoću neuronskih mreža iz bušotinskih i seizmičkih podataka - primjer u naftno-plinskom polju Kloštar : diplomski rad
• Title (english): Structural mapping of seismic and well data with artificial neural networks – an example from oil and gas filed Kloštar : master's thesis
• Author: Filip Vukičević
• Mentor: Josipa Velić (mentor)
• Mentor: Marko Cvetković (pomoć pri izradi, neposredni voditelj)
• Committee member: Josipa Velić (predsjednik povjerenstva)
• Committee member: Tomislav Malvić (član povjerenstva)
• Committee member: Ivan Dragičević (član povjerenstva)
• Granter: University of Zagreb; Faculty of Mining, Geology and Petroleum Engineering; (Department of Geology and Geological Engineering); Zagreb
• Defense date and country: 2013-02-28, Croatia
• Scientific / art field, discipline and subdiscipline: NATURAL SCIENCES; Geology; Geology and Paleontology

Abstract

Područje istraživanja obuhvatilo je naftno-plinsko polje Kloštar tj. dio 3-D seizmičke kocke "Sava Zapad" koji mu pripada. U uvodnim poglavljima prikazani su osnovni podaci polja koji obuhvaćaju zemljopisni položaj, povijest istraživanja, geološke odnose, temeljne postavke neuronskih mreža, povijest njihovog razvoja te opisa metoda korištenih prilikom izradbe ovog rada. Prethodno obradbi podataka neuronskim mrežama, načinje su strukturne karte u vremenskom mjerilu. One su zajedno s prostornim smjestištem EK repera Tg i EK markera Rs7 i Rs5 korištene kao ulaz za uvježbavanje višeslojne neuronske mreže s postupkom povratne informacije. Krajnji rezultat bile su strukturne karte apsolutne dubine. Jedan od ciljeva rada bila je usporedba dubinskih strukturnih karata dobivenih neuronskim mrežama s kartama istog područja ranijem načinjenim metodom kriginga. Metoda neuronske mreže se pokazala uspješnom zbog relativno malog iznosa pogrješke kod predviđanja koja je omogućila dobro definiranje strukturnih odnosa, posebice u dijelovima s manje bušotina.

Abstract (english)

Abstract: The area of research belongs to the oil and gas field Kloštar, more precisely to the seismic 3D cube „Sava Zapad“ crop that associated with it. Basic information of the geographic location, previously made research and the fundamentals of artificial neural networks are described in detail. Firstly the subsurface maps in time domain were made from the available seismic data. Later, these were used as inputs, along with the positions of E-log border Tg and E-log markers Rs7 and Rs5, for neural network analysis. Multy Layer Perceptron neural network was used as a mapping algorithm and the final mapping output were structural maps in absolute depth – true vertical depth subsea. Furthermore, one of the goals of this thesis was the comparison of previously made structural maps with the ordinary krignig method of the same area with those made in this Thesis which shows that neural network approach defined the subsurface structures better in the area with poor well coverage. This proved the neural network method successful for this application.

• Keywords: umjetna neuronska mreža; strukturne karte; polje Kloštar; Hrvatska
• Keywords (english): artificial neural network; structural maps; Kloštar field; Croatia
• Language: croatian
• URN:NBN: urn:nbn:hr:169:549401
• Study programme: Title: Geology; specializations in: Geology of mineral resources and geophysical explorations; Course: Geology of mineral resources and geophysical explorations; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra geologije (magistar/magistra geologije)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Access restricted to students and staff of home institution
• Created on: 2016-01-19 10:49:42