Vilnius Tech DSpace Repository
Not a member yet
52562 research outputs found
Sort by
Analysis and comparison of the effectiveness of renewable energy forecasting in European Union countries
No full textEnergy security is a key element of national security, and its role is becoming increasingly important in the
face of contemporary challenges. The increasing interest in this issue leads to the search for solutions that can contribute
to its strengthening, including the use of renewable energy sources (RES). The aim of the article is to analyze
the possibilities of effective forecasting of energy production from renewable energy sources in the European Union
countries, with particular emphasis on Poland and Lithuania, based on available data. The work will be carried out in
two stages: the first stage will include a statistical analysis of variables related to renewable energy production in the
EU, and the second stage will focus on assessing the effectiveness of forecasting. The data used in the study will come
from publicly available sources, such as Eurostat.Taip / Ye
Assessment of the road safety of the main roads according to the level of severity of injure to road users
No full textLietuvos keliuose žuvusiųjų skaičius per pastarąjį dešimtmetį reikšmingai sumažėjo, tačiau sunkiai ir lengvai
sužeistų eismo dalyvių skaičius nemažėja. Europos Sąjungos strategijoje numatyta iki 2030 m. perpus sumažinti
tiek žuvusiųjų, tiek sunkiai sužeistų asmenų skaičių. Pasirašydama Valetos deklaraciją, Lietuva įsipareigojo skirti daugiau
dėmesio pėsčiųjų ir dviratininkų infrastruktūrai judumo planuose, gerinti eismo dalyvių saugumą plėtojant saugesnę
kelių infrastruktūrą bei eismo įvykyje sunkius sužeidimus patyrusius eismo dalyvius klasifikuoti pagal MAIS3+
sužeidimų klasifikaciją. Atlikto tyrimo metu buvo analizuojami Lietuvos policijos ir Higienos instituto kaupiami duomenys
apie eismo įvykius ir nukentėjusiuosius, vertinamas sužeidimų lygis ir jų pasiskirstymas magistraliniuose keliuose.
Tyrimo metu atlikta avaringumo duomenų analizė, identifikuoti jų trūkumai, apsunkinantys MAIS3+ taikymą. Pateiktos
rekomendacijos, kaip kurti saugesnę kelių infrastruktūrą, ir siūlomi sprendimai, leidžiantys efektyviau vertinti
eismo saugumą ne tik pagal žūčių, bet ir pagal sunkių sužeidimų skaičių.The number of road fatalities in Lithuania has decreased
significantly in the past decade, but the number of
lightly injured road users is also very high. The European
Union’s strategy aims to halve the number of fatalities and
injuries by 2030. By signing the Valletta Declaration, Lithuania
committed itself not only to prioritizing pedestrians and
cyclists in infrastructure plans but also to improving road
safety by developing more road infrastructure. Additionally, it
agreed to classify road users who have suffered serious injuries
in a road accident using the MAIS3+ injury classification. The
study examined data on road accidents and victims collected
by the Police and Hygiene Institute, and analyzed the severity
of injuries and their distribution on main roads. The study
analyzed accident data, identified their shortcomings to implement
MAIS3+, provided recommendations for the development
of safer road infrastructure, and proposed solutions
that allow for an effective assessment of safety not only based
on the number of fatalities, but also based on the number of
serious injuries.Taip / Ye
Analysis of investment strategies in cryptocurrencies
No full textCryptocurrency investment is a rapidly growing financial sector, marked by high volatility, decentralized
technologies, and significant profit potential. Investors use strategies like long-term holding (“HODLing”), portfolio
diversification, and short-term trading. “HODLing” relies on long-term value appreciation but requires resilience to
price fluctuations. Diversifying with assets like Bitcoin and Ethereum reduces risk due to their low correlation with
traditional investments. The crypto market is highly sensitive to geopolitical, economic, and technological factors, attracting
investors during economic instability. Advanced models like LASSO and AutoEncoder aid in price prediction
and strategy optimization. Despite high return potential, careful risk management is essential due to volatility and
regulatory uncertainty. This study experimentally applies identical cryptocurrency portfolios to different investment
strategies, identifying the most profitable approach.Viktoras ChadyšasTaip / Ye
Gyvuliniais nemaistiniais riebalais ir biodegalų mišiniais varomo slėginio uždegimo variklio energinių ir ekologinių rodiklių tyrimas
No full textThe dissertation investigates the energy and environmental performance of a compression ignition engine operating on biodiesel blends derived from animal-based non-food fats and first- and second-generation biodiesel fuels.
The research encompasses experimental and numerical analyses, evaluating how diverse biodiesel compositions influence engine operation. Nine fuel blends, including hydrotreated vegetable oils (HVO) and fatty acid methyl esters (FAME), were prepared and tested under controlled conditions. The research examined key parameters such as ignition delay, combustion pressure, temperature variations, CO₂, CO, NOₓ, HC emissions, and particulate matter (PM). A numerical model using AVL BOOST was developed to simulate in-cylinder combustion processes, providing insights into pressure rise rates, heat release dynamics, and fuel efficiency under various load conditions. The results highlighted the impact of biodiesel composition on combustion phases, revealing that HVO-based blends exhibit shorter ignition delays, improved combustion efficiency, and lower emissions compared to FAME-rich blends. The findings of this dissertation indicate that HVO-based biodiesel blends offer the most promising alternative to diesel, maintaining high energy efficiency while significantly reducing CO, HC, and particulate emissions. However, FAME-rich blends require further optimisation due to their higher viscosity, increased NOₓ emissions, and greater fuel consumption.
The dissertation consists of an introduction, four chapters, a summary of findings, references, and a list of scientific publications by the author. The First Chapter provides an overview of biodiesel properties and their relevance in the transport sector. The Second Chapter details the experimental methodology, including biodiesel preparation, engine testing, and data acquisition techniques. The Third Chapter presents numerical and experimental results, evaluating combustion characteristics, energy indicators, and emission trends. The Fourth Chapter applies machine learning models to predict fuel performance and optimise biodiesel blends. The dissertation concludes with a discussion of the feasibility of biodiesel use in diesel engines and recommendations for future research. This research represents a significant step towards integrating alternative biofuels into the transportation sector, supporting global efforts to reduce greenhouse gas emissions and enhance energy sustainability.
Five scientific articles have been published on the dissertation topic: two in the Web of Science database with a citation index, two in the Web of Science database, conference proceedings, one in other international databases and two in other peer-reviewed scientific journals. Two papers were presented at conferences in Lithuania and one in Poland.Disertacija nagrinėja energinius ir ekologinius rodiklius suspaudimo uždegimo variklyje, veikiančiame su biodegalų mišiniais, pagamintais iš gyvulinės kilmės nemaistinių riebalų ir pirmosios bei antrosios kartos biodyzelino.
Tyrime taikyti eksperimentiniai ir skaitiniai metodai, nagrinėjantys, kaip skirtingos biodegalų sudėtys veikia variklio darbą. Buvo paruošti ir išbandyti devyni degalų mišiniai, įskaitant hidroapdorotą augalinį aliejų (HVO) ir riebalų rūgščių metilo esterius (FAME). Eksperimentuose analizuoti pagrindiniai parametrai, tokie kaip uždegimo vėlavimas, degimo slėgis, temperatūros pokyčiai ir CO₂, CO, NOₓ, HC bei kietųjų dalelių emisijos. Norint detaliau įvertinti degimo procesus, buvo sukurtas skaitinis modelis naudojant AVL BOOST programinę įrangą. Rezultatai parodė, kad HVO pagrindu sudaryti biodegalų mišiniai pasižymi trumpesniu uždegimo vėlavimu, didesniu degimo efektyvumu ir mažesnėmis emisijomis nei FAME turintys mišiniai. Tyrimas parodė, kad HVO pagrindu pagaminti biodegalai yra perspektyviausias dyzelino pakaitalas, nes jie leidžia išlaikyti didelį energijos konversijos efektyvumą ir reikšmingai sumažinti CO, HC ir kietųjų dalelių emisijas. Tačiau FAME turintys degalai reikalauja papildomos optimizacijos, nes dėl jų didesnio klampumo stebimas didėjantis NOₓ emisijų lygis ir didesnės degalų sąnaudos. formavime, siekiant palaikyti perėjimą prie ekologiškesnių transporto sprendimų.
Disertaciją sudaro įvadas, keturi skyriai, rezultatų santrauka, literatūros sąrašas ir autoriaus mokslinių publikacijų sąrašas. Pirmame skyriuje pateikiama biodegalų savybių apžvalga ir jų svarba transporto sektoriuje. Antrame skyriuje išsamiai aprašoma eksperimentinė metodika, įskaitant biodegalų gamybą, variklio bandymus ir duomenų rinkimo procesus. Trečias skyrius pateikia skaitinių ir eksperimentinių rezultatų analizę, įvertinant degimo procesų efektyvumą ir emisijų pokyčius. Ketvirtame skyriuje pristatomi mašininio mokymosi modeliai, skirti numatyti biodegalų poveikį variklio parametrams ir optimizuoti kuro mišinius. Disertacijos išvadose pateikiamos rekomendacijos dėl biodegalų pritaikymo ir tolimesnių tyrimų kryptys. Atliktas tyrimas yra svarbus žingsnis siekiant integruoti alternatyvius biodegalus į transporto sektorių, prisidedant prie pasaulinių pastangų mažinti šiltnamio efektą sukeliančių dujų emisijas ir skatinti energetinį tvarumą.
Disertacijos tema buvo publikuoti penki moksliniai straipsniai: du Web of Science duomenų bazėje su citavimo indeksu, du – Web of Science duomenų bazėje, konferencijų medžiagose, vienas – kitose tarptautinėse duomenų bazėse ir du – kitose recenzuojamuose mokslo žurnaluose. Du pranešimai buvo pristatyti konferencijose Lietuvoje, o vienas – Lenkijoje
Evaluation of MoViNet Streaming Models for Real-Time Action Recognition in Thermal Domain
No full textThis paper investigates the potential of Mobile Video Networks (MoViNet) for real-time human action recognition in the thermal domain. Although MoViNet models have demonstrated strong performance on RGB-based video datasets, their effectiveness on thermal imagery, known for its robustness to low lighting, occlusions, and privacy concerns, remains underexplored. To address this gap, we evaluated three MoViNet variants (A0, A1, A2) using a custom single-person thermal video dataset consisting of three action classes. Due to the limited size of the custom dataset, we apply fine-tuning, GMM-based normalization, and channel replication to adapt thermal inputs. Data augmentation techniques, including brightness adjustments, contrast enhancement, and spatial flips, are used to improve generalization. The findings show that MoViNet A2-stream achieves the highest accuracy (88.33%), with A0 and A1 also showing competitive performance. Real-time visualizations confirm early convergence and high confidence throughout each clip. These findings demonstrate that MoViNet models can be effectively fine-tuned for thermal action recognition with minimal modifications, offering promising potential for real-time deployment in resource-constrained or low-visibility environments.Taip / Ye
Seminaras tyrimų statistikos klausimais. 2025 m.
No full textThe 28th event of the Baltic-Nordic-Ukrainian Network on Survey Statistics and the 6th workshop in Lithuania. Many participants responded to the invited workshop theme “Addressing nonresponse in survey sampling” and prepared their presentation on itInternational Association of Survey StatisticiansLithuanian Statistical AssociationThe Nordplus ProgrammeState Data Agency, Statistics LithuaniaVilnius Gediminas Technical UniversityLithuanian Mathematical SocietyVilnius Universit
Žiedinių apribojimų vertinimas Europos Sąjungos saulės fotovoltikos tiekimo grandinėje
No full textThe dissertation focuses on analysing circularity constraints within a supply chain. The supply chain in this dissertation is defined as a circular supply chain (CSC) that includes materials, components, and end-of-life management of products. The identified research gap is a missing method for organisations to constantly improve circularity in the supply chain. The dissertation aims to propose a circular supply chain theoretical framework model for the continuous improvement of circularity in the CSC. Furthermore, a tool consisting of the sequence of the procedure for identification and assessment of circularity constraints is proposed and applied in the EU photo-voltaics (PV) industry.
The literature analysis in the first chapter resulted in a list of indicators to measure circularity and the proposition of a CSC theoretical framework. The second chapter proposes a tool for identifying and assessing circularity constraints. It also describes the methods used. The third part of the dissertation provides the results of the methods used. Dissertation tasks include (1) analysing CE principles in supply chains, (2) analysing theories and concepts related to CE in supply chains, (3) developing an indicator system to measure circularity in the supply chain, (4) proposing a circular supply chain theoretical framework model, (5) proposing and applying a tool for circularity constraint assessment and solution identification in the EU PV supply chain. The dissertation proposed a circular supply chain theoretical framework model that expands the theory of constraints, knowledge management, resource-based view and resource-dependency theories. The research findings reveal how CE-related knowledge, resources, and capabilities impact the EU PV supply chain’s circularity potential.
The dissertation findings are a list of circularity constraints within the EU PV supply chain. The constraints range from a fragmented regulatory approach for end-of-life PV module management in each EU member state to industry difficulties in cooperating for circularity purposes. The top priority is the hesitant exchange of information. Where most photovoltaics industry components are produced outside the European Union, it is difficult to obtain information regarding materials content, recyclability, disassembly, testing history, etc. The hesitant exchange of information also affects the top four constraints: insufficient traceability of photovoltaic modules, poor logistics handling in the reverse supply chain and illegal end-of-life photovoltaic module exports to non-European countries. The recommendations of the dissertation results suggest synchronising the regulatory landscape across the EU member states, standardising end-of-life PV module operators’ data sharing for circularity purposes, and implementing a centralised database in the EU to manage all CSC data related to circularity
Biblioteka informuoja, 2025 Nr. 17 (714)
No full textNaujai į Web of Science ir Scopus įtrauktų Vilnius Gedimino technikos darbuotojų publikacijų sąrašai ir kitos bibliotekos aktualijos.17 (714)202
Navigating innovation partnerships: challenges and opportunities in startup-corporate collaborations
No full textCorporations actively pursue startups for rapid innovation, while startups seek to enhance their business models with larger companies’ support, creating an ideal partnership. However, the initiation of such relationships poses challenges for both parties. The study clarifies interaction pathways among entities and how partnerships create new technology evaluation models. The research used a qualitative methodology with content analysis techniques involving 18 interviews conducted with founders of Brazilian startups. By elucidating the real impact of such collaborations, the study offers valuable insights for practitioners and researchers seeking to understand and leverage the dynamics of startup-corporation relationships for mutual benefit and innovation.Taip / Yes