16168 research outputs found
Sort by
An integrated stacked convolutional neural network and the levy flight-based grasshopper optimization algorithm for predicting heart disease /
No full textCardiovascular disease is the leading cause of death worldwide, including critical conditions such as blood vessel blockage, heart failure, and stroke. Accurate and early prediction of heart disease remains a significant challenge due to the complexity of symptoms and the variability of contributing factors. This study proposes a novel hybrid model integrating a Stacked Convolutional Neural Network (SCNN) with the Levy Flight-based Grasshopper Optimization Algorithm (LFGOA) to address this challenge. The SCNN provides robust feature extraction, while LFGOA enhances the model by optimizing hyperparameters, improving classification accuracy, and reducing overfitting. The proposed approach is evaluated using four publicly available heart disease datasets, each representing diverse clinical and demographic features. Compared to traditional classifiers, including Regression Trees, Support Vector Machine, Logistic Regression, K-Nearest Neighbors, and standard Neural Networks, the SCNN-LFGOA consistently outperforms these methods. The results highlight that the SCNN-LFGOA achieves an average accuracy of 99%, with significant improvements in specificity, sensitivity, and F1-Score, showcasing its adaptability and robustness across datasets. This study highlights the SCNN-LFGOA's potential as a transformative tool for early and accurate heart disease prediction, contributing to improved patient outcomes and more efficient healthcare resource utilization. By combining deep learning with an advanced optimization technique, this research introduces a scalable and effective solution to a critical healthcare problem
Innovations in digital technology for virtualizing fashion products.
No full textIn the final thesis the process of fashion product virtualization and its benefits for creators and consumers are reviewed. Different methods of virtualization using computer-aided design programs, 3D scanning tools, and generative artificial intelligence are analyzed. The use of virtual fashion products in video games and wearing virtual clothes in photos and using AR filters is reviewed. Cases of digital fashion shows and virtual fashion stores are analyzed. Then, the use of virtualization to promote and preserve cultural heritage is reviewed. In the research part, a dress is virtualized using 6 artificial intelligence tools, one computer program and 5 mobile photogrammetry applications, and the virtual design program CLO3D. The generated results are analyzed and the advantages, disadvantages and efficiency of the programs are presented. When comparing the 3D models, recommendations are prepared on where they can be used best. Based on the results and recommendations, women's lithuanian national costume is virtualized. Using the photogrammetry program Reality Capture and the virtual design program Blender, a static virtual costume model was created, which was animated to turn around. The created model visualizations can be used in ecommerce and in the virtualization of ethnographic museums. Using the specialized virtual design program CLO3D, costume visualizations and animations were created. Generative artificial intelligence was used to create accessories that can be used in games as environment decorations or additional parts of a virtual costume
Research into underwater concreting technologies.
No full textThis master’s thesis is based on an experimental study of underwater concreting technology. The study investigated ways to find an alternative to the special admixture used for underwater concreting, replacing it with a conventional stabilizing admixture that increases the stability and cohesion of the concrete mix. The main objective was to determine the impact of a stabilizing admixture on the slump of the mixture and setting of concrete underwater and its influence on concrete’s physical and mechanical properties. The thesis was based on literature analysis and experimental studies. The literature analysis describes the field of concrete application, its advantages and disadvantages, the properties of the concrete mix, the methods used for placing the mix underwater, the physical and mechanical properties of the concrete, and an overview of the admixtures used. The experimental study includes reviewing the materials used in the research and the preparation of cementitious doughs and concrete mixes with different compositions. Tests were carried out on the spread ability of the cement dough and the flexibility of the concrete mix, the determination of the air content of the mix, the washout of fines from the concrete mix, and the formation of specimens in tap water, lagoon water and seawater. Determination of density, compressive strength and water penetration depth after hardening the concrete specimens. After the above tests, the results of the tests were evaluated
Research of pinned joints of steel structures reconstruction into semi-rigid joints.
No full textThe final Master's project analyses the reconstruction of pinned nodes into semi-rigid using finite element and component methods. The study analyses the influence of the reconstructed semi-rigid nodes on the individual elements and the global stability of the steel frame. The reconstruction of steel hinge connections into semi-rigid connections increases the overall stability and performance of the structures. The reconstruction of hinged connections into semirigid connections results in a more uniform distribution of stresses in the components and a reduction of structural displacements and overall deformations of the steel structures. The aim of this paper is to find and identify the most effective methods for reconstructing steel hinge nodes into semi-rigid joints by means of an analytical analysis of the literature and a global steel frame, investigating the behavior of these nodes and reducing the beam utilization of the additional solar electric load. Also to determine the combination of the reconstructed semi-rigid nodes that achieves the lowest utilization of the steel beams in terms of safety and serviceability limit states, taking into account the additional load of the solar plant. The study evaluates the components that have the highest positive effect and reduce the utilization of steel structures and increase the torsional stiffness value of the node by reconstructing the hinged connections into semi-rigid ones. With a better understanding of the possibilities and advantages of reconstructing hinged joints into semi-rigid joints in a more optimal way, researchers can contribute to the improvement of design and practice guidelines for structural engineering. Preliminary determination of the effect of connections on a framed structure, where many variables need to be taken into account, is time-consuming and demanding. Therefore, the most rational solutions for the reconstruction of hinged connections into semi-rigid connections are analyzed and presented in this study. This will facilitate and speed up the iterative design, fabrication and erection process of structures. The project consists of the following parts as specified in the assignment: introduction, literature review, methodology, analytical studies, conclusions, list of references, summary in English and three chapters. The introduction discusses the relevance of the work, the research problem and the object of the study. The aim and objectives of the work are formulated, the methodology of the study is presented, and the scientific novelty and practical usefulness of the work are noted. The conclusions of the analytical calculations using the finite element and component method are presented
Mutual trust assessment of leaders and educators in educational institutions.
No full textModern educational institutions face various challenges such as constant changes, increasing competition, integration and preventive processes, and decreasing human resources. This complex environment imposes requirements on educational institutions and demands new solutions, often related to effective leadership, as leaders are responsible for creating a trustful atmosphere, which is essential for better student achievements, higher teacher job satisfaction, and overall school performance. Trust is a complex concept that encompasses honesty, openness, reliability, competence, and goodwill. A trusting atmosphere fosters positive activity within the school community, encourages curiosity, motivation to learn, and professional development. In contrast, a lack of trust creates tension, decreases employee motivation, and hinders both student and teacher progress. The object of the study is the mutual trust between leaders and teachers in educational institutions. The aim of the study is to analyze mutual trust between leaders and teachers in educational institutions in the city of Panevėžys. The objectives to achieve the aim are: to examine the theoretical foundations of mutual trust between leaders and employees (the concept of trust, dimensions of trust in leaders and trust in employees), to analyze research on mutual trust between educational institution leaders and employees, and to empirically evaluate mutual trust between leaders and employees in educational institutions. The methodology chosen to achieve the research objectives includes analysis of scientific literature, interviews, written surveys, qualitative (content) analysis, and descriptive statistics. The analysis of scientific sources suggests that trust is a key factor influencing organizational success, employee satisfaction, and work atmosphere. However, it is quite difficult to define it unambiguously, as different perspectives on trust include moral commitment, communication and collaboration, mutual expectations, and relationships with people and the social environment. These perspectives help understand the multifaceted nature of trust and its importance both in personal and professional contexts, encouraging strong and effective social bonds and long-term cooperation. Promoting trust in educational institutions helps create better working conditions for teachers, who can freely share ideas, develop professionally, and aim for higher student achievements. Therefore, trust in leaders is one of the key factors contributing to the creation of such an environment. The analysis of scientific literature helped identify the key dimensions of trust in leaders (managerial competence, communication, honesty, goodwill, reliability, and openness) and trust in employees (honesty and fairness, employee loyalty, creativity, psychological well-being, employee productivity, and effectiveness) in educational institutions. The empirical part of the study, based on quantitative and qualitative research results, indicates that the mutual trust between both leaders and teachers has a significant impact on the work environment and organizational success. The quantitative data show that teachers positively assess the managerial qualities of their leaders; however, certain aspects, such as decision-making transparency and communication, still raise concerns. The qualitative study revealed that teachers value honesty, openness, and emotional support, which strengthen their trust in the leader and loyalty to the organization. Trust and collaboration between both parties promote creativity, independence, and improve work results. It is likely that the research results will contribute to the improvement of educational institutions' performance, and the practical recommendations presented in the study will help improve teaching quality, reduce tension, and create a more open and motivating environment, where all members of the community trust each other and can achieve better results
Youth perception of public spaces in the city /
No full textFor any urban development, the opinion of the final stakeholders is essential. However, the uncensored youth perception regarding the urban environment is rarely included in the decision-making process. Here, we present the youth’s perception of public spaces in the city. The empirical research was conducted in Kaunas, which is a typical tertiary city. That is why such results are comparable with findings from other middle-sized cities around the world. The photovoice research method was applied to gather data on youth perception of public spaces. The dataset consisted of 793 unique responses from young people. The qualitative analysis was conducted by using MaxQDA 24 software. Our findings suggest that young people prefer to spend time with friends in man-made urban environments rather than in natural or semi-natural green spaces. Many of their favourite places are considered to be well suited to young people’s needs. Suggestions for improving their favourite places mostly revolve around built infrastructure, yet many of them emphasise the importance of sustainable urban development
Robust forest sound classification using Pareto-Mordukhovich optimized MFCC in environmental monitoring /
No full textAs a complex ecosystem composed of flora and fauna, the forest has always been vulnerable to threats. Previous researchers utilized environmental audio collections, such as the ESC-50 and UrbanSound8k datasets, as proximate representatives of sounds potentially present in forests. This study focuses on the application of deep learning models for forest sound classification as an effort to establish an early threats detection system. The research evaluates the performance of several pre-trained deep learning models, including MobileNet, GoogleNet, and ResNet, on the limited FSC22 dataset, which consists of 2,025 forest sound recordings classified into 27 categories. To improve classification capabilities, the study introduces a hybrid model that combines neural network (CNN) with a Bidirectional Long- Short-Term Memory (BiLSTM) layer, designed to capture both spatial and temporal features of the sound data. The research also employs Pareto-Mordukhovich-optimized Mel Frequency Cepstral Coefficients (MFCC) for feature extraction, improving the representation of audio signals. Data augmentation and dimensionality reduction techniques were also explored to assess their impact on model performance. The results indicate that the proposed hybrid CNN-BiLSTM model significantly improved classification loss and accuracy scores compared to the standalone pre-trained models. GoogleNet, with an added BiLSTM layer and augmented data, achieved an average reduced loss score of 0.7209 and average accuracy of 0.7852, demonstrating its potential to classify forest sounds. Improvements in loss score and classification performance highlight the potential of hybrid models in environmental sound analysis, particularly in scenarios with limited data availability
In situ analysis of magnesium ([NH4]2Mg[CO3]2·4H2O), copper ([NH3]2Cu(CO3)), and zinc ((NH3)Zn(CO3)) ammonium carbonate thermal properties /
No full textThree metal ammonium (ammonia) carbonate double salts were mechanochemically synthesized including magnesium ammonium carbonate ([NH4]2Mg[CO3]2·4H2O, MAC), copper ammonium carbonate ([NH3]2Cu(CO3), CAC), and zinc ammonium carbonate ((NH3)Zn(CO3), ZAC), and their crystallinity and thermal stability were investigated using in situ X-ray diffraction. The crystal structures were investigated in the temperature interval from 25 to 355 °C. MAC exhibited relatively low thermal stability with its crystal structure rapidly losing crystalline water as well as ammonium ions already at 85 °C and transforming into an amorphous carbonate, as confirmed by in situ infrared spectroscopy. CAC and ZAC exhibited loss of the corresponding NH3 at much higher temperatures transforming into a mixture of the corresponding metal oxides according to the XRD with the outer layer of the carbonate, as suggested by infrared spectroscopy measurements. The thermogravimetric analysis has been carried out to investigate the thermal degradation behavior. Initial fast mass loss was observed in the parent [NH4]2CO3 already under 100 °C with complex mass loss patterns from basic carbonate precursors achieving stable mass at above 500 °C for magnesium carbonate. CAC and ZAC inherited intermediate responses to temperature when compared to the precursors exhibiting stable mass at ~ 300 °C. This can be associated with the distinct crystal structure of the compounds whereby CAC and ZAC exhibit strong bonds with metal ions, while in MAC crystalline water contributes to lower stability of the crystal as well as the loosely bound ammonium ion. Infrared spectra were obtained and interpreted at room temperature followed by their in situ thermal analysis up to 250 °C in attenuated total reflection mode. Changes in the complex room temperature spectra with increasing temperature were interpreted as loss of N–H bonds, as confirmed by the concurrent decrease in the 3350–3000 and 1248 cm−1 band region. It was found that ZAC was the most thermally stable compound among the three double salts. The thermal data obtained in this work have practical implications for nutrient recovery and their release in the environment
Research on the use of renewable energy resources in the heating system of railway switches.
No full textThis master's thesis analyses the potential for integrating renewable energy sources (RES) into railway switch heating systems, aiming to contribute to the implementation of strategic projects developed by LTG Group, which will significantly aid in mitigating climate change and support LTG Group’s goal of becoming a climate-neutral organization by 2050. The relevance of this research is also driven by the need to ensure the uninterrupted operation of railway transport under adverse weather conditions, while reducing energy consumption and greenhouse gas (GHG) emissions. One of the areas within LTG's infrastructure where energy consumption optimization would be particularly impactful is the railway switch heating systems. The main objectives of this thesis are to examine the possibilities of integrating RES into switch heating systems to enhance energy efficiency and reduce dependence on fossil fuels, to assess the preconditions and effectiveness of RES application, to develop an experimental research methodology, and to perform empirical calculations under real conditions. The theoretical part discusses the classification of railway switches, their structural elements, and operational issues, particularly those related to snow and ice accumulation during the winter season. A comprehensive analysis of switch heating technologies is provided, covering compressed air, gas, resistive electric, inductive, and geothermal heating methods. Each system is evaluated in terms of its energy conversion efficiency, operational reliability, cost-effectiveness, and technical applicability. Inductive heating technology, operating at frequencies of 40–70 kHz, demonstrates greater energy efficiency compared to resistive heating, whereas geothermal systems, although economically viable in warmer regions, are limited under extremely low temperatures. The overview of RES presents global data on the use of solar, wind, hydro, biomass, and geothermal energy sources. Technologies of Horizontal-Axis Wind Turbines (HAWT) and Vertical-Axis Wind Turbines (VAWT) are analysed, including the influence of turbine axis orientation on efficiency, site selection criteria, and operational characteristics. The possibilities for photovoltaic module installation, their conversion efficiency, and economic indicators are also discussed. Data from various sources indicate that wind and solar energy are the fastest-growing sectors within renewable energy. The empirical part covers research conducted at Pilviškiai railway station, where parameters of switch heating systems across different switch groups — temperature, electricity consumption, and the influence of weather conditions — were measured under real-world conditions. Certified measuring equipment was used, including FLIR ThermaCam I7 thermal imaging cameras and EMSItest IR– 8839 infrared temperature sensors. Experiments were carried out on five switch groups (VS21–VS25) at the Pilviškiai railway yard, recording temperature and electricity consumption data across different switch operation cycles. The collected data were statistically analyzed, with the calculation of average consumption, its dependence on environmental conditions, and identification of optimization opportunities for the heating system. Weather condition analysis included examining variations in wind speed and solar radiation fluxes, which enabled the determination of RES implementation efficiency limits within the Lithuanian climate context. The study also analysed and evaluated five different wind turbine manufacturers, assessing their performance, installation requirements, and suitability for the specifics of the Pilviškiai site. Additionally, a site survey of the station area was conducted to identify the optimal location for turbine installation, taking into account landscape topography, infrastructure constraints, and access to energy flows. In summary, the integration of RES into switch heating infrastructure is technically feasible, economically viable, and environmentally beneficial. However, the final choice of technologies must be based on a comprehensive assessment of site characteristics, weather conditions, and energy consumption profiles
Research of high aspect ratio wings aeroelastic behavior at low Reynolds numbers.
No full textThe final project researches the problematics of aeroelastic phenomena at low Reynolds numbers, with experiments related to high-aspect-ratio wings characterized by low mass, thickness and chord length, but large wingspan. Such characteristics result in increased sensitivity to the onset of aeroelastic phenomena, as the wing is highly flexible and prone to bending and torsion under lower airflow loads. In the literature review part, the problematics of aeroelasticity are analyzed, identifying the main types of static and dynamic phenomena, with the main focus on flutter analysis. Key requirements for wing design, modeling and experimental research are also discussed. An overview of the applied methodologies and computational models are provided. In the final project, wing models of different geometries are designed, numerical analysis is carried out to validate geometric characteristics and to determine possible deformation scenarios. Using physical models, the main elastic parameters are calculated: bending and torsional stiffness, which are necessary for the numerical analysis of aeroelasticity phenomena, in order to find instability speeds using the K and P-K methods. The results show that increasing the number of composite layers and decreasing chord length lead to higher airflow speeds required to reach flutter and divergence. Therefore, the most reliable models identified in this project are wings with 2 to 3 layers and chord lengths of 0,02 m to 0,03 m, as their instability speeds met the defined critical airflow speed criteria. For experimental analysis, wing-tip vibration measurement equipment is designed and manufactured. During the experiments, wing-tip vibrations in a wind tunnel are studied using the developed measurement equipment and a laser accelerometer. Vibration and frequency analyses are performed, the experimental results are compared with each other as well as with the numerically obtained instability speed values. The validity of the experiments is supported by similar critical airflow speeds and frequency characteristics indicating the onset of instability. Experiments show that the difference between the flutter speeds obtained from K and P-K methods and experimental data is 0–4%. Divergence speed discrepancies are larger, ranging from 2% to 20%, due to differences between theoretical and actual models geometries and stiffness characteristics. The consistency of results demonstrated that both experimental methods are suitable for investigating aeroelastic phenomena under low Reynolds number conditions, additionally, their combination allows even more reliable evaluation of the dynamic response of the structure and verification of numerical model accuracy. The main objective of the project is to research the aeroelastic phenomena of high-aspect-ratio wings at low Reynolds numbers by integrating the fields of aerodynamics, structural mechanics, composite materials, electronics and data analysis through the application of numerical and experimental methods