Iskanje po katalogu digitalne knjižnice

Opis: V okviru diplomskega dela je bil izdelan hladilni sistem za električni ojačevalnik, ki ga uporabljamo za napajanje piezoelektričnih aktuatorjev. Preračunali smo temperature spoja ojačevalnika, izbrali ustrezen hladilnik in določili ustrezne ventilatorje za aktivno hlajenje sistema. Električni ojačevalnik smo nato testirali pri vhodnih signalih različnih frekvenc in amplitud, ob tem pa smo opazovali izhodni signal in temperaturo ohišja močnostnega ojačevalnika PB51. S programom Apex Power Design smo opravili analizo stabilnosti sestavljenega električnega ojačevalnika. Za lažje in bolj kvalitetno zajemanje fotografij in posnetkov smo 3d natisnili tudi nastavek za mobilni telefon, ki se pritrdi na okular mikroskopa. Tako lahko sedaj zajemamo počasne posnetke kroženja vode na mikroskopskem stekelcu, kar nam bo olajšalo nadaljnje raziskave na področju mikrofluidnih motorjev.
Ključne besede: Močnostni ojačevalnik, piezoelektrični aktuator, hladilni sistem, 3d tiskanje, SolidWorks, Apex Power Design.
Objavljeno v DKUM: 19.09.2022; Ogledov: 114; Prenosov: 14
Celotno besedilo (4,05 MB)

Opis: In the diploma thesis the flow pattern and conditions in the downcomer of a pressurized water nuclear reactor were simulated, to study coolant mixing inside the primary circuit. This was done using computational dynamics software Ansys CFX. The nuclear reactor in Krško (Slovenia) was used as a basis for the model. The downcomer geometry was simplied for a more feasible study. Three separate simulations were done: one with normal conditions, one with loss of coolant accident conditions and one with a modified geometry. The flow was then studied at different condition at the same levels using velocity contours. The results showed that the axial velocity profiles at the downcomer outlet are not completely uniform.
Ključne besede: Nuclear Power Plant, Pressurized Water Reactor, Downcomer, Flow simulation, Ansys CFX
Objavljeno v DKUM: 13.10.2021; Ogledov: 508; Prenosov: 47
Celotno besedilo (2,49 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: 809; Prenosov: 133
Celotno besedilo (13,35 MB)

Opis: Ko se lotimo študiranja vsebin o poslovnem obveščanju, si težko predstavljamo, da veliko podjetij posluje brez primernih poslovnih orodji oziroma na svoj prilagojen način. V naši diplomski nalogi bomo najprej razdelali pojem poslovno obveščanje. Kaj pomeni, kaj ga sestavlja in kaj se z njim lahko počne. Glede na temo naše naloge se bomo kasneje dotaknili tudi malih in srednjih podjetji, v nadaljevanju s kratico MSP. Predstavili bomo predvsem to, kako jih med seboj ločimo, kakšne so njihove posebnosti ter ugotavljali, kakšne smernice bi MSP morala upoštevati pri izbiri orodja za poslovno obveščanje. Na koncu jih bomo primerjali tudi s tem, kar ponuja orodje Power BI. Microsoft je na področju poslovnega obveščanja dolga leta vztrajal z dodatki k orodju Excel, vendar s tem komaj sledil konkurenci. Kot odgovor na to, je Microsoft razvil povsem novo orodje, Power BI, ki ga je v tem segmentu poslovanja ponovno povzdignil v sam vrh. Čeprav sta na trgu prisotni tudi dve plačljivi verziji orodja, nam že brezplačna verzija ponuja neverjetne možnosti upravljanja s podatki, pri čemer je omejitev le uporabnikova domišljija.
Ključne besede: Microsoft, MSP, poslovno obveščanje, Power BI
Objavljeno v DKUM: 17.11.2020; Ogledov: 678; Prenosov: 62
Celotno besedilo (1,25 MB)

Opis: In this master’s thesis, we have established a model of a transformer winding. This model is based on the multi transmission line theory that allows us to model each turn of the winding separately. By using that approach, it was possible to calculate the frequency response of a winding in the required frequency range from 20 Hz to 2 MHz. The purpose of this simulation was to evaluate the frequency response analysis as a diagnostic tool for deformations of windings in power transformers. An evaluation was performed using sensitivity analysis of several parameters that indicate winding deformations. Finally, axial bending deformation was simulated and successfully diagnosed by applying the frequency response analysis.
Ključne besede: Power transformer, Frequency response analysis, Diagnostics, MTL model
Objavljeno v DKUM: 03.11.2020; Ogledov: 511; Prenosov: 55
Celotno besedilo (2,64 MB)