Primjena strojnog učenja u energetskim sustavima

undergraduate thesis

• Title (english): Machine learning application in power system
• Author: Magda Smolić-Ročak
• Mentor: Bojan Trkulja (mentor)
• Committee member: Bojan Trkulja (predsjednik povjerenstva)
• Committee member: Ana Drandić (član povjerenstva)
• Committee member: Stjepan Frljić (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2022-08-29, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Tema ovog rada jest primjena strojnog učenja u energetskim sustavima realizirana programskim jezikom Python. Sustavi koji se bave proizvodnjom energije nastoje ostvariti optimalnu energetsku učinkovitost. S tim ciljem, rad se bavi korištenjem različitih algoritama i modela strojnog učenja u svrhu predviđanja potrošnji energije. Na ulaz modela se dovode informacije o temperaturama i količinama vlažnosti, a na izlazu model generira predviđenu potrošnju energije. Jednostavna regresija i višestruka linearna regresija su korištene za izračun predviđene kumulativne satne i dnevne potrošnje energije. Korištenjem neuronske mreže, izgrađen je model za izračun i prikaz aproskimirane funkcije satnih potrošnji energije tijekom jednog dana. Uz dobivene informacije o budućim potrošnjama, elektrane mogu unaprijed pretpostaviti koliku količinu energije će generirati u odabrano vrijeme.

Abstract (english)

The topic of this paper is the application of machine learning in energy systems. Systems that deal with energy production strive to achieve optimal energy efficiency. To this end, the paper deals with the use of different algorithms and models of machine learning for the purpose of predicting energy consumption. At the input of the model, information on temperatures and amounts of humidity is brought, and at the output, the model generates the predicted energy consumption. Simple regression and multiple linear regression were used to calculate the predicted cumulative hourly and daily energy consumption. Using the neural network, a model was built to calculate and display the approximated function of hourly energy consumption over one day. With the obtained information on future consumption, power plants can predict in advance how much energy they will generate at the selected time.

• Keywords: strojno učenje; jednostavna regresija; višestruka linearna regresija; aproksimacija funkcije
• Keywords (english): machine learning; simple regression; multiple linear regression; function approximation
• Language: croatian
• URN:NBN: urn:nbn:hr:168:799376
• Study programme: Title: Computing; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-01-15 11:13:13