Repository of FERIT Osijek
Not a member yet
4053 research outputs found
Sort by
Influence of the bus type selection on power flow calculations
No full textU ovom radu je glavno razmatranje bila procjena utjecaja odabira različitih tipova čvorova na
rezultate proračuna tokova snaga. Provođenje navedenog razmatranja je započelo postavljanjem
pitanja o mogućnosti utjecaja promjena tipova čvorova na napone sabirnica, ukupne tokove snaga
te ukupne gubitke djelatne snage u mreži kroz rješavanje proračuna tokova snaga za jednostavni
model mreže od tri sabirnice. Nakon čega je predstavljen relevantniji model za takva razmatranja
u vidu IEEE 14 sabirničke mreže u kojem su se vršile promjene tipova sabirnica. Promjene tipova
čvorova, odnosno sabirnica, postignute su odspajanjem određenih sinkronih strojeva te
konfiguriranjem sabirnica u Power World Simulatoru. Promjene tipova sabirnica su provedene
kroz 14 simulacijskih slučajeva podijeljenih u 2 simulacijska seta. Predstavljeni su i interpretirani
rezultati s obzirom na koje su se izveli zaključci o zanemarivosti utjecaja odabira različitih tipova
čvorova na promjene iznosa napona na sabirnicama i primjetnom utjecaju na ukupne gubitke
djelatne snage s obzirom na model koji je bio u razmatranju.This paper primarily focused on assessing the influence of selecting different types of buses on
power flow calculations. The examination began by questioning the potential impact of bus type
changes on bus voltages, overall power flows, and total losses in the network, achieved through
power flow calculations for a simple three-bus network model. Subsequently, a more relevant
model, namely the IEEE 14-bus system, was introduced over which bus type changes have been
made. These modifications in bus types were accomplished by disconnecting specific synchronous
machines and configuring the buses in Power World Simulator. Bus type changes have been
implemented across 14 cases divided into two simulation sets. The results presented led to the
conclusion that the impact of selecting different bus types on voltage variations at buses was
negligible. At the same time, a noticeable influence has been observed on the total losses
concerning the presented model
Measuring sugar levels with a Hanna Instruments 96801 digital refractometer
No full textCilj rada je razviti algoritam za prepoznavanje znakova sedam segmentnog pokaznika sa zaslona
digitalnog refraktometra. Slika se dohvaća pomoću kamere koja ju šalje putem WiFi veze.
Obradom slike dolazi se do parametara koji se predaju modelu koji pomoću modela nasumiĉne
šume prepoznaje o kojim se znakovima radi. Rješenje je napravljeno u programskom jeziku
Python, ĉime je osigurano da se program moţe pokrenuti u više razvojnih okruţenja te je time
program i dostupniji.The goal of this work is to develop an algorithm for character recognition to be used with a
seven-segment display of a digital refractometer. An image is acquired using a digital camera
that sends it using a wireless connection. Using digital image processing the algorithm obtains
parameters which are then passed to a random forest machine learning model which recognizes
the displayed characters. The solution was written using Python programming language, which
ensures the program can be run in multiple IDEs, making the program widely available
Comparison of relational and NoSQL databases
No full textOkosnica završnog rada je usporedba relacijskih i NoSQL baza podataka. Kroz
objašnjena strukture te prednosti i nedostataka pojedine baze podataka i njihove skalabilnost te
izradom i objašnjenjem izrade oba tipa baza podataka oni su uspoređeni po njihovom načinju
skaliranja te po njihovoj strukturi. U cilju dubljeg razumijevanja razlike između oba tipa baza
podataka odrađena je usporedba nad izrađenim primjerima gdje se odrađivala usporedba nad
unosom podataka, brisanju podataka te ažuriranju podataka kao i usporedba nad obradom
određenih upita za svaku od baza podataka.
Rezultat je sveobuhvatan pregled razlika i prednosti korištenja različitih vrsta baza
podataka, pružajući vrijedan izvor za istraživače i stručnjake na tom području.The backbone of the final paper is a comparison of relational and NoSQL databases.
Through the explained structures and the advantages and disadvantages of individual databases
and their scalability, and by creating and explaining the creation of both types of databases,
they were compared according to their scaling method and their structure. In order to gain a
deeper understanding of the difference between both types of databases, a comparison was
made of the created examples where a comparison was made of data entry and the speed of
retrieving the same data.
The result is a comprehensive overview of the differences and advantages of using
different types of databases, providing a valuable resource for researchers and practitioners in
the field
The possibility of implementing a variable frequency transformer in a real power system
No full textProblematika prijenosa velikih djelatnih i jalovih snaga, ali i samog pogona dugačkih prijenosnih
vodova najviših napona kroz zamkaste i radijalne mreže na velike udaljenosti postaje danas vrlo
interesantna i zanimljiva. S jedne strane javljaju se problemi kutne i naponske stabilnosti vodova,
pa onda i posljedično EES-a, koji se mogu i približiti opasnom granicama stabilnosti uslijed čega
može doći do naponskog sloma. Posebni važno pitanje predstavlja utvrđivanje realnog iznosa
gubitaka i ekonomičnosti takvog prijenosa snaga u slučaju pogona dugačkih prijenosnih vodova
najviših napona. Zahtjevi za prijenosom velikih snaga iz udaljenih proizvođačkih područja u
velika potrošačka područja, nameću potrebu ispitivanja mogućnosti primjene varijabilnih
frekvencijskih transformatora za povezivanje tih udaljenih područja. Cilj ovog rada je optimirani
parametre sustava i dati prikaz na realnoj mreži i EES-u.The problem of transmission of large active and reactive forces, but also the operation of long
transmission lines of the highest voltages through trapped and radial networks over long distances
is becoming very interesting and interesting today. On the one hand, there are problems of angular
and voltage stability of lines, and then consequently of the power system, which can approach the
dangerous limits of stability, which can lead to voltage breakdown. A particularly important issue
is to determine the real amount of losses and the cost-effectiveness of such power transmission in
the case of operation of long transmission lines of the highest voltages. The requirements for the
transmission of high power from remote production areas to large consumer areas, impose the
need to examine the possibility of using variable frequency transformers to connect these remote
areas. The aim of this paper is to optimize the system parameters and give an overview of the real
network and the power system
Steganography - practical examples
No full textU ovom su diplomskom radu navedene i opisane različite tehnike steganografije te su prikazani
različiti praktični primjeri. Opisana su osnovna načela steganografije, povijest steganografije,
podjela steganografije i detaljno je objašnjen princip rada steganografskih tehnika. Navedene
su i opisane tehnike steganalize. Nadalje, navedeni su mogući načini primjene steganografskih
tehnika, kao i neki od primjera tih primjena iz stvarnog života. Za prikaz praktičnih primjera
korišteni su različiti besplatni steganografski alati koji se temelje na nekoj od steganografskih
tehnika. Za svaki od primjera opisan je postupak i način korištenja alata, a dobiveni rezultati
prikazani su slikama.In this master thesis, various steganography techniques are listed and described along with a
few practical examples. The basic principles of steganography, the history of steganography,
the division of steganography and the working principle of steganographic techniques are
explained in detail. Steganalysis techniques are listed and described. Furthermore, possible
ways of applying steganographic techniques are listed, as well as some examples of these
applications from real life. Various free steganographic tools based on some of the
steganographic techniques were used to show practical examples. For each of the examples,
the procedure and method of using the tool is described, and the obtained results are shown in
pictures. Conclusion is given on various techniques and their application in real life
Measurement of losses in magnetic sheets using the Epstein method
No full textU završnom radu opisane su metode ispitivanja magnetskih svojstava materijala te je provedeno
mjerenje gubitaka u magnetskim limovima metodom Epsteina. Povijest o magnetskim
materijalima ukratko je opisana, kao i nastanak i razvoj magnetskih materijala. Nadalje su opisana
svojstava magnetskih materijala te su navedene vrste ponašanja magnetskih materijala koji se
nalaze u magnetskom polju. Također uz njih opisane su i pojave koje se javljaju pri magnetiziranju
poput petlje histereze te sami gubitci prilikom magnetiziranja. Prije provedbe mjerenja potrebno
je upoznati se s vrstama mekih i tvrdih magnetskih materijala kao i s vrstama mjerenja iz kojih se
dobivaju mjerni rezultati pomoću kojih se mogu izračunati gubitci. Pri provođenju mjerenja
metodom Epsteina dobiveni su mjerni rezultati iz kojih su izračunati gubitci koji su prikazani
tablično i grafički te su objašnjene različite vrijednosti gubitaka korištenih uzoraka magnetskih
materijala. Također je provedeno snimanje petlje histereze pomoću osciloskopa te su prikazane
petlje histereze za spomenute uzorke.The final thesis describes methods for testing the magnetic properties of materials, and
measurements of losses in magnetic laminations were conducted using the Epstein method. The
history of magnetic materials is briefly described, including their origins and development. The
properties of magnetic materials are described, and the types of behaviors exhibited by magnetic
materials in a magnetic field are listed. The phenomena that occur during magnetization, such as
the hysteresis loop, as well as the losses during magnetization, are also explained. Prior to
conducting the measurements using the Epstein method, it is necessary to familiarize oneself with
the types of soft and hard magnetic materials, as well as the types of measurements from which
measurement results are obtained for calculating losses. By performing measurements using the
Epstein method, measurement results were obtained, from which losses were calculated and
presented in tabular and graphical form. The different loss values of the magnetic material samples
used are explained. Hysteresis loop recording using an oscilloscope was also performed, and
hysteresis loops for the mentioned samples are presented
The use of reactive components in the transmission system
No full textU ovom završnom radu opisana je regulacija napona u prijenosnom sustavu, kako elementima
samog sustava tako i regulacijskim elementima koji se ugrađuju ukoliko sustav nije u mogućnosti
sam otkloniti promjene. Prikazani su konvencionalni i moderni načini regulacije te je dana opaska
na trenutno ugrađene elemente u prijenosnom sustavu Republike Hrvatske.In this thesis the voltage regulation in transmission system is described, with the elements of the
transmission system and regulation systems which are installed when the transmission system
itself can't resolve those changes. Conventional and modern ways of voltage regulation are shown,
and there is a comment of regulation systems installed in the Republic of Croatia transmission
system
Floating Photovoltaic Systems
No full textCilj ovoga rada je opis najnovije tehnologije fotonaponskih sustava, plutajući fotonaponski
sustavi (PFN). Opisuje se trenutno stanja razvoja PFN-a i veličina takvih sustava te se navode
zemlje sa najvećim instaliranim kapacitetom. Promatra se prednost hibridizacije PFN sustava sa
hidroenergijom te se istražuje utjecaj PFN sustava na kvalitetu vodenog ekosustava. Također se
razmatraju troškovi implementacije PFN sustava te iznose troškovi po svakoj komponenti
sustava. Na samom kraju iznosi se izračun nivelirane cijene električne energije (LCOE) u
ovisnosti o različitim parametrima sustava: očekivanoj godišnjoj proizvodnji električne energije,
diskontnoj stopi sustava, stopi inflacije te stopi degradacije sustava.
Ključne riječi: plutajući fotonaponski sustav (PFN), razvoj, troškoviThe aim of this paper is to describe the latest technology of photovoltaic systems, floating
photovoltaic systems (PFN). The current state of PFN development and the size of such systems
are described, and the countries with the largest installed capacity are listed. The advantage of
hybridizing the PFN system with hydropower is observed and the impact of the PFN system on
the quality of the water ecosystem is investigated. The costs of the implementation of the PFN
system are also considered, as well as the costs for each component of the system. At the very
end, the calculation of the levelized cost of electricity (LCOE) is presented depending on various
system parameters: expected annual electricity production, system discount rate, inflation rate
and system degradation rate
WWVB transmiter
No full textCilj ovog rada bila je izrada odašiljača kojim bi se pomoću WWVB standarda podešavale ure koje
podržavaju taj standard. Za izradu je korištena Arduino platforma zbog mnoštva mogućnosti koje
ona pruža. Wemos D1 mini Lite odabran je zbog mogućnosti povezivanja na Wi-Fi mreže putem
kojih se preuzima informacija o vremenu. SparkFun Pro Micro odabran je zbog svog stabilnog
ugrađenog oscilatora koji ima mogućnost generiranja frekvencije od 60 kHz koja je potrebna za
WWVB standard. Kao pojačalo korišteno je LM358 opracijsko pojačalo, a na njega je spojena
antena izrađena od feromagnetne jezgre i žice namotane preko nje. Korištenjem odgovarajućih
biblioteka i algoritama ostvarena je potrebna programska podrška.Goal of this paper was to create a transmitter which would use WWVB standard to set clocks
which support that standard. The Arduino platform was used for the creation due to the large
number of possibilities. Wemos D1 mini Lite was chosen due to its possibility to connect to WiFi networks through which information about the time is obtained. SparkFun Pro Micro was
chosen because of its stable built-in oscillator which has the possibility to generate 60 kHz
frequency required by WWVB standard. LM358 was used as an operational amplifier to which
the antenna is connected. Antenna was made using a ferromagnetic core and a large number of
windings. Required software support was accomplished using appropriate libraries and algorithms
Facial emotion recognition
No full textEmocionalno stanje čovjeka može se prepoznati iz izraza čovjekova lica. Iako čovjek može doživjeti
puno različitih emocija, sve se one mogu svrstati u neke od glavnih kao što su ljutnja, gađenje, strah,
radost, tuga, iznenađenost te neutralnost. Problem prepoznavanja emocija sa ljudskog lica u
računarstvu moguće je riješiti uporabom strojnog učenja gdje će istrenirani model za sliku čovjekova
lica izreći emociju koja prevladava. Ovo područje je zanimljivo zato što može poboljšati interakciju
čovjeka i računala. U isto vrijeme je i vrlo izazovno zbog različitih uvjeta u kojima se uzorkuju slike
na kojima je potrebno prepoznati emociju, sličnosti između nekih emocija te subjektivnog dojma koji
kod ljudi utječe na prepoznavanje emocija. U ovome radu reproducirano je pet postojećih state-ofthe-art rješenja koja se temelje na konvolucijskoj neuronskoj mreži. Najbolje rješenje korišteno je za
konačnu aplikaciju koja prepoznaje emocije sa ljudskog lica u stvarnom vremenu. Uspješnost
pojedinih rješenja evaluirana je na FER-2013 skupu podataka.The emotional state of a person can be recognized from the expression on their face. Even though a
person can experience a wide range of different emotions, these emotions can be classified into several
main classes such as anger, disgust, fear, joy, sadness, surprise, and neutrality. The problem of
recognizing emotions from human face in computer science can be solved by using machine learning,
where a trained model for facial images can identify the predominant emotion in a person's picture.
This field is intriguing because it can enhance human-computer interaction, but at the same time, it is
quite challenging due to the various conditions under which images are sampled for emotion
recognition, the similarities between some emotions, and the subjective impression that affects
human's recognition of emotions. In this master thesis, five existing state-of-the-art solutions based
on convolutional neural networks were reproduced, and the best solution was used for the final
application that recognizes emotions from human face in real-time. The performance of each solution
was evaluated using the FER-2013 dataset