Iskanje po katalogu digitalne knjižnice

Opis: V diplomskem delu smo izdelali programsko opremo, ki bi z uporabo kompresijskega postopka brez izgub zmanjšala velikost datotek in minimizirala čas pošiljanja oz. naredila sistem v vesolju bolj odziven. Naredili smo splošni pregled algoritmov za kompresijo in za glavni algoritem izbrali CCSDS 121.0-B-2 ter ga primerjali s sorodnima algoritmoma LZW in Heatshrink. Testiranje smo izvedli nad sivinskimi slikami in binarnem zapisu, ki smo ga zajeli iz GPS-a. Najbolj primeren algoritem za stiskanje sivinskih slik je CCSDS 121.0-B-2, zaradi kompresijskega razmerja in časa kompresije. Najbolj primeren algoritem za stiskanje krajših binarnih zapisov pa Heatshrink, zaradi kompresijskega razmerja in izredno kratkega časa dekompresije.
Ključne besede: kompresija, dekompresija, vgrajeni sistem, Skylabs, PicoSky™ evaluacijska plošča, CCSDS 121.0-B-2, Heatshrink, LZW, slike, splošni binarni zapisi
Objavljeno v DKUM: 18.10.2021; Ogledov: 192; Prenosov: 29
Celotno besedilo (1,18 MB)

Opis: V magistrskem delu je opisan razvoj vgrajenega sistema za večkanalni zajem analognih veličin, katerega naloga je merjenje električne napetosti na več vzporednih kanalih z velikim dinamičnim območjem. Po korakih je opisano načrtovanje strojne opreme, vgrajene programske opreme za krmiljenje programirljive logike in programske opreme računalniškega vmesnika. Podrobno so predstavljene in analizirane meritve opravljene na prototipnem izdelku, s katerimi je overjeno delovanje končnega produkta.
Ključne besede: merilni sistem, analogno-digitalna pretvorba, FPGA, vgrajen sistem
Objavljeno v DKUM: 18.10.2021; Ogledov: 190; Prenosov: 20
Celotno besedilo (2,68 MB)

Opis: The increasing demand for smaller, faster, and more efficient switch mode power supplies (SMPS) and energy converters has led to the development of new power MOSFET technologies, such as SiC (silicon carbide) power MOSFET transistors. Switching parameters, output voltages and currents and power consumption need to be measured accurately to assess different types of power electronics components more efficiently in today’s rapid implementation of new energy transfer and modulation technology. The need for improving switching circuits in power electronics and trying to use simple simulation circuits for designing PCBs to measure output values of said power electronics transistors is the basis of this thesis. Even though manufacturers of power electronics transistors provide data on drain-source voltages and currents, these are mostly measured for specific conditions regarding pulse duration, gate-source voltage, case temperature of the transistor etc., and as such do not provide information as to where the limits of the drain source voltages and currents for different measurement parameter values are. The goal of the thesis is to find a way to measure the limits of drain-source voltages and currents of power electronics transistors (in this case a SiC MOSFET transistor), and to see if a simple SPICE-model could be built for other power electronic transistors as well, based on the results of this thesis.
Ključne besede: SiC, MOSFET, transistor, simulation, PCB
Objavljeno v DKUM: 02.06.2021; Ogledov: 349; Prenosov: 35
Celotno besedilo (4,35 MB)

Opis: Human-Elephant Conflicts are a major problem in terms of elephant conservation. According to WILDLABS, an average of 400 people and 100 elephants are killed every year in India alone because of them. Early warning systems replace the role of human watchers and warn local communities of nearby, potentially life threatening, elephants, thus minimising the Human-Elephant Conflicts. In this Master's thesis we present the structure of an early warning system, which consists of several low-power embedded systems equipped with thermal cameras and a single gateway. To detect elephants from captured thermal images we used Machine Learning methods, specifically Convolutional Neural Networks. The main focus of this thesis was the design, implementation and evaluation of Machine Learning models running on microcontrollers under low-power conditions. We designed and trained several accurate image classification models, optimised them for on-device deployment and compared them against models trained with commercial software in terms of accuracy, inference speed and size. While writing firmware, we ported a part of the TensorFlow library and created our own build system, suitable for the libopencm3 platform. We also implemented reporting of inference results over the LoRaWAN network and described a possible server-size solution. We finally a constructed fully functional embedded system from various development and evaluation boards, and evaluated its performance in terms of power consumption. We show that embedded systems with Machine Learning capabilities are a viable solution to many real life problems.
Ključne besede: machine learning, microcontroller, on-device inference, thermal camera, low-power system
Objavljeno v DKUM: 06.01.2021; Ogledov: 459; Prenosov: 89
Celotno besedilo (13,35 MB)

Opis: V magistrskem delu je opisan razvoj sistema za natančno določanje položaja in orientacije. Sistem temelji na nizkocenovnih in lahko dostopnih komponentah. Za določanje položaja smo uporabili sprejemnik GNSS, ki ga za izboljšanje določitve položaja s pomočjo kalmanovega filtra, združujemo z inercijsko merilno enoto. Inercijsko merilno enoto sestavljajo pospeškometer, žiroskop in magnetometer. S pomočjo žiroskopa in magnetometra določamo orientacijo merilne enote, na ta način pretvorimo pospeške izmerjene s pospeškometrom v koordinatni sistem sprejemnika GNSS. V delu je predstavljeno delovanje sistema GNSS in Kalmanovega filtra. Hkrati so v delu zajeti različni koordinatni sistemi, ki jih potrebujemo za pravilno delovanje sistema. Opisano je tudi delovanje vgrajene programske opreme in njen razvoj. V zadnjem delu magisterskega dela so predstavljene meritve in delovanje merilnega sistema.
Ključne besede: GNSS, inercijska merilna enota, Kalmanov filter, vgrajen sistem
Objavljeno v DKUM: 03.11.2020; Ogledov: 348; Prenosov: 60
Celotno besedilo (3,95 MB)