National Repository of Grey Literature
Not a member yet
574846 research outputs found
Sort by
Two-Generation Family House
No full textThe subject of the bachelor's thesis is the creation of construction documentation for a twogeneration family house. The house is designed for a family of five and a family of two. The family house is located on a slightly sloping plot. The building is partially basement-based and has two above-ground floors. The building is based on basic passports. The above-ground vertical perimeter structures are designed with Porotherm system, the underground with reinforced concrete. The horizontal load-bearing structures are designed as monolithic reinforced concrete. The roofing of the building is solved by a combination of a flat green roof and a truss gable roof. The drawing part was created with Archicad
Export and display of QoS statistics
No full textTato práce se zaměřuje na vytvoření řešení pro sběr, ukládání a zobrazení statistik QoS ze síťových zařízení. Řešení využívá databázi VictoriaMetrics pro ukládání dat a pro vizualizaci nástroj Grafana. V práci je použit i nástroj Authentik pro autorizaci a autentizaci uživatele. Celý systém je nasaditelný pomocí nástroje Docker a je připraven na práci s více teoretickými zákazníky najednou. V řešení byly použity pouze open-source verze nástrojů
Design and manufacture of a casting of a smooth Cambridge cylinder ring
No full textThe subject of this thesis is the design and production of a smooth ring casting for a Cambridge-type agricultural roller. The main objective is to evaluate the possibility of a design modification of the original component, with an emphasis on increasing its service life. The original one-piece design was revised into a two-piece solution featuring a replaceable hub insert, allowing for reduced costs in the event of wear in the central area. As part of the experimental production, models were created using 3D printing, moulds were manufactured and castings were made for two material variants – GJL and GJS. The selected materials, manufacturing technologies, and the resulting castings are described and compared in terms of their properties. The findings confirm the technical feasibility of the proposed solution and highlight its benefits in terms of maintenance economics and operational efficiency
Automatic classification of breast ultrasound images
No full textThis bachelor’s thesis deals with the automatic classification of breast ultrasound images using machine learning methods, with a focus on deep neural networks. The introduction of the thesis provides a detailed description of the anatomy and pathology of the breast, including benign and malignant tumors. Subsequently, various diagnostic methods used in the screening and diagnosis of breast diseases, such as mammography, magnetic resonance imaging, and ultrasound, are presented, highlighting their advantages and disadvantages. The thesis also includes an overview of current solutions and techniques used in the field of automatic classification of ultrasound images, including a review of publicly available databases and their parameters. Based on the literature review, appropriate deep learning models are selected, implemented, and subjected to training
Study of a Joint Path for Pedestrians and Cyclists near Blatnice pod Svatým Antonínkem
No full textThe bachelor's thesis in the form of a study proposes a variant solution for a shared path for cyclists and pedestrians. The section is located between Veselí nad Moravou and Blatnice pod Svatým Antonínkem. The shared path continues the existing cycle path parallel with road I/54. The study considers the ownership relationships in the proposed route and the cost of the project. The path partly uses existing purpose roads. Four variants were developed, and one was selected as the most suitable
Developement and testing of outdoor package for cleaning of surface waters based on wheat bran biochar
No full textAccess to safe drinking water remains a challenge, particularly in remote regions and during outdoor activities, where technical infrastructure for water treatment is often unavailable. This bachelor thesis therefore focuses on designing a practical solution for treating water from surface sources, using biochar produced from wheat bran as an accessible and environmentally friendly adsorbent material, in accordance with the principles of the circular economy. The theoretical part addresses the issue of chemical and microbiological water contaminants, the principles of adsorption, commonly used adsorbents, legislative requirements for drinking water, and also reviews existing drinking water treatment solutions. The experimental part evaluates the adsorption capabilities of variously modified wheat bran biochars in comparison with commercial activated carbon on model contaminants such as selected toxic metals, malachite green, and Escherichia coli bacteria. Additionally, adsorption efficiency was tested on collected river water. The results show that wheat bran biochar exhibits promising properties for use in field conditions and may serve as part of alternatives to conventional methods
Technological Solution of a Flat Roof of an Apartment Building in Bzenec
No full textThe subject of this bachelor thesis is the implementation of the roofing phase of a new residential building in the town of Bzenec. The proposed structure is a flat biosolar roof combining a photovoltaic system with extensive greenery. The thesis focuses on the technological aspects of constructing this roofing system, including material transport, construction methodology, bill of quantities, cost estimate, and site layout design. It also includes a construction schedule, machinery setup proposal, inspection and test plan, and solutions related to occupational health and safety. Alternative technological approaches are presented, and detailed construction drawings and technical details are provided in the appendices
Cycling trip planning application
No full textThe use of modern information technology is increasingly being applied to areas such as sport, including cycling. However, despite the increasing availability of digital maps and planning tools, users often lack detailed information on the difficulty of sporting activities, especially on the climbing sections that represent the most performance-intensive parts of routes. The aim of this thesis is to provide detailed statistics and graphical analyses of planned cycling trips. The proposed application should enable more efficient planning and support systematic preparation for adequate allocation of physical forces. The main benefit of the solution should be to facilitate the process of choosing an appropriate strategy to successfully overcome the set challenges not only for amateur enthusiasts but also for professionals. The outputs of the application in the form of elevation profiles should serve primarily to gain a deeper understanding of the exact course of trips and hilly passages
Predictive Maintenance of Embedded Systems Based on Machine Learning
No full textPrediktivní údržba (PdM) využívá metody založené na datech k předvídání poruch zařízení a ke snižování neplánovaných odstávek. S nástupem technologie Tiny Machine Learning (TinyML) je nyní možné provozovat lehké prediktivní modely přímo na vestavěných mikrokontrolérech, což umožňuje decentralizovanou a okamžitou detekci poruch na okraji sítě. Tato práce zkoumá proveditelnost a efektivitu použití TinyML v PdM na výpočetně omezeném hardwaru prostřednictvím dvou případových studií. První případ využívá dataset Microsoft Azure PdM k natrénování více-třídního klasifikátoru poruch pomocí spektrálních příznaků a hustých neuronových sítí. Výsledný model dosáhl 65,5% přesnosti na validační sadě a 52,2% na testovací sadě, přičemž F1 skóre pro jednotlivé třídy se pohybovalo mezi 0,20 až 0,40. Simulace ukázaly, že falešně pozitivní a falešně negativní detekce omezují přínosy modelu a přinášejí pouze mírné zlepšení dostupnosti zařízení. Druhý případ se zabývá detekcí anomálií v reálném čase na bezkomutátorovém stejnosměrném motoru (BLDC) pomocí vývojové desky Arduino Nano 33 BLE Sense. Binární klasifikátor dosáhl přesnosti 89%, přesnosti a citlivosti 0,82 a AUC 0,92. Ačkoliv se nasazení modelu do zařízení nezdařilo kvůli problémům s integrací SDK, výkonové metriky ze systému Edge Impulse potvrdily proveditelnost inferencí v reálném čase (latence 25ms, spotřeba paměti 16KB RAM). Simulované strategie údržby ukázaly, že dobře fungující modely mohou výrazně zvýšit provozuschopnost a snížit počet poruch, zatímco nedostatečně natrénované modely nemusejí přinášet téměř žádný přínos. Energetický profil navíc potvrdil, že inference pomocí TinyML spotřebovala 4–5× méně energie než přenos dat do cloudu, což potvrzuje její vhodnost pro efektivní PdM přímo na zařízení
Design and stress-strain analysis of meta-material structures with gradient cell wall thickness
No full textThis thesis focuses on the creation and stress-strain analysis of auxetic metamaterials with variable cell wall thickness. The introductory part provides a brief overview of mechanical metamaterials, with an emphasis on their unusual properties. The methodology and a reliable procedure for creating computational models are then described, including their export to ANSYS Mechanical APDL, where all simulations are performed. The computational models are gradually created using SolidWorks, nTopology, and ANSYS SpaceClaim. The practical part analyzes two 2D structures – a reentrant honeycomb and a double arrowhead structure, each designed in three variants of gradient cell wall thickness. The final part of the thesis presents the results of the analysis, especially the force reaction of the structures and stress distribution in critical areas and demonstrates how geometric modifications influence the mechanical response of the structures