Modeli Tranzicije i Predviđanja Sistema Snabdevanja Prirodnim Gasom

Abstract

Kroz definisani predmet istraživanja disertacija je dala širi okvir tranzicije snabdevanja i planiranja sistema prirodnog gasa. U prvom delu prikazano je stanje globalne energetike uz detaljnu analizu regionalnih tržišta. Takođe, analizirana su dešavanja koja se tiču energetike Republike Srbije. U datom okviru, istražen je istorijat transformacije svetske i regionalne energetike sa fokusom na tržišta prirodnog gasa. Definisani su ključni trendovi tranzicije energetskih tržišta na osnovu kojih su projektovani budući scenariji i evolucija sektora prirodnog gasa. Ključni zaključci predstavljeni su za kratkoročnu, srednjoročnu i dugoročnu tranziciju sektora prirodnog gasa u Republici Srbiji. U drugom delu disertacije razvijeni su inovativni modeli na bazi veštačke inteligencije za potrebe kratkoročnog predviđanja cene i potrošnje prirodnog gasa kao univerzalnih i opšte primenljivih algoritama predikcije. Formirani modeli procene daju ključan doprinos u budućoj trgovini gasom, ali i u procesima proizvodnje, skladištenja i distribucije gasa. Ovakvim pristupom se takođe omogućava održivo ekonomsko planiranje i zaštita životne sredine. Kao dodatak predstavljeni modeli doprinose i boljem planiranju proizvodnje drugih vidova energije.

Alternate abstract:

The research objective of this Ph.D. thesis is to provide a general framework for natural gas distribution and planning systems. The first part of the thesis introduces current state of global energetics along with a detailed analysis of regional markets. The analysis also includes recent energy developments in the Republic of Serbia. Specifically, the research examines the historical context of global and regional energetics with an emphasis on natural gas markets. It identifies key energy transition trends based on which future scenarios and developments in the natural gas sector are projected. Several conclusions are drawn that concern short-term, medium-term, and long-term transitions in the natural gas sector of the Republic of Serbia. The second part of the thesis conceptualizes innovative artificial intelligence models for the purpose of very short-run prediction of natural gas price and consumption. By their nature, these models represent universal and widely applicable prediction algorithms. Such models offer a wealth of information that can be used in future natural gas trading, as well as in production, storage, and distribution of natural gas. By design, the models from this thesis also incorporate sustainable economic planning and environmental protection processes. In addition, natural gas models presented in this work effectively contribute to improved production planning of alternative energy sources.