Modeliranje električne potrošnje korisnika pomoću IoT uređaja i neuronskih mreža

master's thesis

• Title (english): Modelling of user power consumption by IoT controller with neural networks
• Author: Ante Gulin
• Mentor: Vinko Lešić (mentor)
• Committee member: Vinko Lešić (predsjednik povjerenstva)
• Committee member: Mario Vašak (član povjerenstva)
• Committee member: Branimir Novoselnik (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2020-07-15, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Electrical Engineering

Abstract

Zgrade su jedan od najvećih potrošača energije na svijetu i također su prostor za značajne uštede energije. Učenje korisničkog ponašanja preduvjet je za prognoziranje i moguću optimizaciju cjelokupne energetske bilance prilagođene ponašanju korisnika. Da bi se to postiglo potrebno je pratiti ponašanje korisnika. IoT uređaj koristio se za otkrivanje prisutnosti korisnika i njihove potrošnje električne energije na razini od jedne godine, u različito doba dana, vremenske uvjete, godišnja doba i radne / neradne dane. Nakon prikupljanja podataka napravio se matematički model koji se temelji na informaciji te je naučen pomoću neuronskih mreža. Jednom naučeni model implementiran je kao Python skripta na mikroračunalo Rasberry Pi i pruža predviđeno ponašanje korisnicima 36 sati unaprijed.

Abstract (english)

Buildings are one of the largest energy consumers in the world and also a space for significant energy saving. Learning user behavior is a prerequisite for forecasting and possible optimization of the entire buldings energy balance adjusted to user behavior. In order to do so, it is necessary to monitor the behaviors of the users. IoT device was used for detecting the presence of users and their electricity consumption at the level of one year, at different times of the day, weather conditions, seasons and working/non-working days. After dana collection, mathematical model based on infomation and lusing neural networks were made. Once learned model was implemented as a Python script on Rasberry Pi microcomputer and it is delivering predicted user behavior 36 hours in advance.

• Keywords: štednja energije; Internet stvari; neuronska mreža; matematički model; Raspberry Pi; LSTM; RNN; Python; tensorflow; pandas
• Keywords (english): energy saving; IoT; neural network; mathematical model; Raspberry Pi; LSTM; RNN; Python; tensorflow; pandas
• Language: croatian
• URN:NBN: urn:nbn:hr:168:700554
• Study programme: Title: Electrical Engineering and Information Technology; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka elektrotehnike i informacijske tehnologije (magistar/magistra inženjer/inženjerka elektrotehnike i informacijske tehnologije)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-01-13 12:23:17