ABSTRACT. OPENING SPEECH

ABSTRACT. Greeting Speech Representative Rector Board

ABSTRACT. Global warming is threatening process which can not deny any more. The consequence could be fatal both for the nature and for the human society; increasing number of species threatened with extinction; the weather is getting more extreme. The building sector is responsible for about 40% of the carbon emissions including the manufacturing of building materials and the operation of the buildings. To stop this process carbon emissions should be radically reduced which needs innovative materials and technologies. By photosynthesis produced wood, and other lignocellulose materials mostly built up from the CO2 content of the atmosphere must play higher importance even in the near future. The higher the ratio of natural based building materials the higher the environment protection effect of the building. From this reason, the increasing utilization of cellulose based materials such as rubber wood, palm tree banana fiber or any other renewable building materials reduce the harm of the environment. There are new technologies and materials developed for producing effective thermal insulation materials at the University of Sopron Hungary and also in research companies. E.g. Mirrorpanel system by using multiple heat reflections or the binderless rigid insulation materials can reduce the heat transfer of the wall and the newly developed Intelligent Windows System can reduce the heat transfer up to the half of the original windows. A Horizon 2020 project aims at to develop a MiniStor system. By the MiniStor thermal storage system is a promising solution for air conditioning and heating driven by solar energy. Energy storage capacity is more than ten times higher than that of the water storage capacity. Such innovative technologies can help the mankind for sustainable development and for promising future.

ABSTRACT. Nowadays, electrical potential, current, impedance, capacitance, etc. play an important role in our daily life, and these electrical parameters can actually have many applications. For example, electrical impedance spectroscopy (EIS) has been widely used for the characterization of (biological) substances. There are many applications of EIS, and the speaker cited his own research experience in applying EIS in E. coli. Identification and quantification, as well as characterization of microplastics.

In my E. coli. identification and quantification study, a biorecognition-element-free interdigitated microelectrode (IDμE) sensor is designed and developed with good reliability and affordability. Results show that the designed sensor can identify E. coli with good selectivity using an impedance and capacitance of 7.69 MHz. At its optimum impedance of 1.3 kHz, the IDμE sensor can reliably quantify E. coli (Figure 2) in a range of measurement (103.2~106 cfu/mL), linearity (R2 = 0.97), sensitivity (18.15 kΩ/log (cfu/mL)), and limit of detection (103.2 cfu/mL). Therefore, the IDμE sensor developed possesses high potential for industrial and clinical applications.

In my microplastics identification study, EIS measurements using IDμE confirmed the accurate identification of microplastic materials in question, by using self-normalized ratio between two characteristic frequencies of 7 MHz and 8.9 MHz, Z’f=7 MHz/Z’f=8.9 MHz. 3-kNN classifier built with the ratio Z’f=7 MHz/Z’f=8.9 MHz, and Z’f=8 MHz/Z’f=8.9 MHz, demonstrates accuracy upto 90% for the identification of single or both microplastic types in samples (Figure 4). These results confirm impedance spectroscopy, permitting rapid identification of microplastic without labelling and skillful techniques, as a potential rapid sensor.

ABSTRACT. Transparent-flexible thermoelectric thin films have immense potential as power supplies for future small-sized consumer electronics, the internet of things, and wearable devices. Here, we report the thermoelectric properties of dual Ga and In doped ZnO (IGZO) films deposited on a polyimide substrate with post-thermal treatment in vacuum along with fabricating 4-unileg flexible IGZO thermoelectric devices. All the as-deposited and annealed IGZO films are the preferred (002) orientation and under tensile stress. The post-thermal treatment controls the dopant substitution/diffusion in the host ZnO lattice affecting the film crystallinity, residual stress, and thermoelectric properties. Among the films, the IGZO film annealed at 250°C has the best power factor of 16.9 μWm1K2 with the largest crystal size, lowest tensile stress, highest carrier concentration, and lowest density-of-state effective mass. The practical application of flexible IGZO films was also reported via a 4-unileg-IGZO films thermoelectric module, which achieved an output power about 3.2 nW at ∆T = 120 K.

ABSTRACT. The fusion of simulation and data represents a transformative paradigm in the realm of Computer-Aided Engineering (CAE) and Finite Element Analysis (FEA). For decades, simulation has served as a foundational tool for engineers, offering the ability to model and analyze intricate systems with impressive precision. However, conventional simulations have often relied on oversimplified assumptions and theoretical models, potentially overlooking the complexities inherent in real-world phenomena. The advent of abundant data, coupled with advancements in artificial intelligence and data analytics, ushers in a unique opportunity to elevate the authenticity and dependability of simulation outcomes. This synergy of data-driven methodologies with CAE and FEA workflows empowers engineers to harness the wealth of accessible data sources to enhance simulation accuracy and efficiency. Real-time sensor data, experimental findings, and historical performance records can seamlessly integrate into simulations, ushering in dynamic and adaptive models capable of responding to dynamic conditions. This presentation will offer an in-depth exploration of specific applications arising from the convergence of simulation and data within CAE and FEA. By seamlessly infusing data-driven approaches into established simulation processes, engineers can unlock novel insights, optimize designs, and expedite the product development lifecycle. In summation, this presentation seeks to stimulate discussions surrounding the potential and challenges on the horizon for this convergence, cultivating collaborations and propelling innovation within the realm of CAE and FEA.

ABSTRACT. Consumer requirements for the safety and quality of food are becoming a serious problem in both developed and developing countries. Meanwhile, traditional methods do not bring high efficiency. It is the driving force behind the application of new techniques in food safety and quality monitoring. Optical-based methods are a particularly important part of this field because of their non-invasive, rapid, and wide applicability. In work, we design a low-cost model of food quality assessment based on optical imaging technique. The model is integrated with temperature and humidity sensors to monitor the influence of the environment on the survey object. The initial results show the possibility of applying optical imaging techniques in food quality assessment and provide a foundation for further studies in applying available machine learning models to food assessment and classification.

ABSTRACT. This study creates simulations of electric vehicle (EV) traffic based on their expected market share and EV adoption scenario on an access-controlled highway and evaluates capacity utilization to determine the long-term financial viability of the EV charging infrastructure developed along the highway. The model predicts the charging events, the number of chargers required, the amount of electricity required for EV charging, and the capacity utilization of EV charging infrastructure over a certain time period. The outcomes highlight the EV charging business's sensitivity to EV traffic volume and the growing vehicle range. The results will help the EV charging business's proliferation based on the maturity of EV market.

ABSTRACT. The knowledge of the propagation of light in biological tissues plays a particularly important role in research related to the therapeutic application of light (LLLT/PBM) and diagnostic imaging. In our work, we simulated the propagation of laser beam in plane (flat beam) and Gaussian shape at 800 nm into breast tissue to study the influence of beam shape on absorption, scattering and fluence in breast tissue. The simulation scenario is built including the survey of normal breast tissue and breast tissue with malignant tumor (breast cancer) based on the MCMatlab program in three-dimensional form. The results on absorption distribution, diffuse reflection,... show a difference between the normal breast tissue and breast tissue with malignant tumor. In addition, the results show that the incident beam shape has an influence on the absorption distribution, diffuse reflectance in the two types of breast tissue investigated. This study has implications for the development of new methods for breast cancer diagnosis based on near-infrared light.

ABSTRACT. Globalization and technological advancements have contributed a shift towards hyperspecialization--the division of work into more specialized pieces done by multiple people--to achieve improvements in quality, efficiency, and cost in the labor market. However, there is a mismatch between the specialized skills employers are looking for and the available skillsets of unemployed workers. Neuroscience research on long-term persistent training in highly specialized trades and neuroplasticity can bring insight into solutions for these labor market challenges, illustrating how our brains can change with a changing job market. This review looks at 10 experimental studies to see how specific job requirements and cognitive demands influence the brain's structure and function. These studies observed structural and functional neuroplasticity due to long-term persistent training in two domains of skills: athletic/motor and cognitive memory skills. Findings show how intensive training in both motor and cognitive skills instigates remarkable changes in the brain's ability to adapt and evolve, even in adulthood, having implications for workforce policy and future work on occupational neuroscience.

ABSTRACT. Segmenting brain tumors from MRI images is an important task that greatly aids in the diagnosis and treatment of brain tumors. However, manual segmentation is time-consuming and labor-intensive due to the variability in size and location of tumors. Therefore, the paper proposes a deep learning-based method to automatically detect and segment brain tumors from MRI images. The research implements the construction of a U-Net network architecture and suggests using an improved version called Attention U-Net for tumor segmentation from MRI images. The results of the two models, U-Net and Attention U-Net, are compared using the same dataset during the training and testing process. The dataset used in this study consists of nearly 4000 images, which are split into training and testing sets. The results of the Attention U-Net model show an accuracy pixel of 99.7%, an Intersection over Union (IoU) index of 0.780, and a Dice index of 0.873 when performing five-fold cross-validation. The findings demonstrate that the proposed approach achieves high performance and accuracy according to evaluation criteria. An automated and accurate brain tumor segmentation model could greatly support physicians in making diagnostic decisions in the future.

ABSTRACT. Abstract - TBA

ABSTRACT. Abstract — TBA

ABSTRACT. The researching project get ideas from parallel bar that are already on the market and in many hospitals, with the goal of finding a method to solve the product's limitations and add some needed functions. The results of the research are twelve 3D images of parts of the parallel bar. The result of research brings convenience to transportation because the parts can be seperated together. The overall shape of the model when completely assembled is two parallel bars stretched in a closed circle. The overall size of the product can be changed thanks to changing the quantity and length of horizontal connecting bars. The product is convenient to disassemble and store when the patient finishes exercising so as not to take up space. At the same time, the product also integrates a back belt accessory to keep the patient from falling during recovery exercises, improving the patient's fear of falling during exercise and making it safer to let the patient carry out the treatment themselves. The height of the product and the length of the bar connecting the belt and the frame can be adjusted to make the user as comfortable as possible.

ABSTRACT. Humidity control is critical for reserving medicine. Because when the humidity is appropriate (below 50% and over 70%), the efficiency and effectiveness of the medicine are affected; even in some types of medicine, the high humidity environment can cause denaturation and become toxic, endangering the user's health. This project is inspired by existing humidity control systems on the market or in pharmacies, with the goal of lowering the total cost and adding certain important features to the device. The model of humidity control system for medicine storage room is using the ESP32 microcontroller to collect the humidity, temperature data from the DHT22 sensor and to control the exhaust fan. The user can observe the temperature, humidity and fan status thougth a website. There are two ways of operation: automation and manual. When the automated mode is enabled, the exhaust fan will activate when the humidity exceeds 70% and switch off when the humidity falls below 60%. The manual mode allows the user to regulate the exhaust fan via buttons. Furthermore, a website is built to allow the user to monitor and control the system whenever they have an Internet connection.

ABSTRACT. This paper aims to develop an automated stroke detection system, a convolution neural network (CNN) deep learning approach. Computed Tomography (CT) and Magnetic Resonance Imaging (MRI) are the primary imaging modalities for stroke imaging. However, the first has more advantages, like fast imaging is more popular and still presents good-quality visual information. Hence, the data were 3D computed tomography images, including 92 brain strokes and 129 normal brain volumes. To deal with the lack of data, data augmentation was used. The data varied in size, so I used a slice selection method called Spline Interpolated Zoom to make them have the same size. Then, brain volumes were put into a CNN. The model was built with convolution layers, batch normalization, activation functions, and a fully connected layer. The model generates predictions if a brain volume contains areas of stroke or not, with 5-fold cross-validation. After training for about 50 epochs, the model showed its effectiveness as the accuracy improved. The results obtained: the proposed model was overfitting, yet its average accuracy is almost 95%, and the average loss function was 0.417 through 5 folds.

ABSTRACT. The paper investigates a plate made of Functionally Graded Material (FGM) with variable mechanical properties. It focuses on studying the temperature variation and deformation of the plate under constrained thermal expansion and circumferential thermal loading. The obtained results are compared with previous studies to ensure the paper's reliability. The theoretical analysis considers the thin plate behavior of a power-law distributed FGM plate with a constant thickness. Different modeling approaches are employed, including composite plate modeling and ANSYS software simulation. MATLAB software is also used to apply the thin plate theory. Comparisons are made between the results obtained from these approaches and those reported by other researchers, highlighting the strengths and limitations of each method.

ABSTRACT. In this work, we investigate the hydrogen and oxygen evolution reaction (HER/OER) of MnxOy – derived metal-organic framework (Mn-BTC) powder synthesized by solvothermal method. The crystal structure and its HER/OER properties of MnxOy and Mn-BTC were characterized by using PXRD, FT-IR, TGA, N2 isotherm at 77K and Gamry interface 1000 electrochemical workstation by three-electrode configuration system. We found that the pristine Mn-BTC is not a good HER catalyst, however, after calcination at high temperature, the HER activity is drastically enhanced. Obviously, the Mn-MOF-500 electrocatalyst exhibits the best HER activity with overpotentials of 199 mV and 319 mV to afford the current densities of 10 mA cm-2 and 100 mA cm-2, respectively, substantially lower than that those for the Mn-MOF-300 (260 mV and 399 mV), Mn-MOF-700 (249 mV and 355 mV), and Mn-BTC (301 mV and 469 mV). Beside the outstanding HER activity, we found that this Mn-MOF-500 electrocatalyst also shows an excellent electrochemical performance for OER in the same electrolyte. Specifically, the Mn-MOF-500 electrocatalyst shows excellent OER catalytic activity with low overpotentials of 387 and 429 mV to obtain the current densities of 10 and 100 mA cm-2, respectively, superior to that of Mn-MOF-300 (428 and 549 mV), Mn-MOF-700 (412 and 531 mV), and Mn-BTC (458 and 614 mV).

ABSTRACT. Functionally Graded Materials (FGMs) have gained significant attention in recent years due to their unique property distribution and enhanced structural performance. This paper presents a comprehensive study of the dynamic behavior and potential applications of Functionally Graded Materials (FGMs) plates through free vibration analysis. The research begins with a detailed review of FGMs, plate theories, and vibration analysis techniques to establish a solid theoretical foundation. Using Ansys-CAE software, a mathematical model is solved using numerical analysis to determine FGM plates' natural frequencies and mode shapes. The accuracy and reliability of the proposed model are confirmed through comparisons with existing solutions and published research. Parametric studies are conducted to investigate the influence of material gradient, plate dimensions, and boundary conditions on the dynamic response of FGM plates. The results reveal that the material gradient significantly shapes the natural frequencies and mode shapes, allowing for tailored dynamic behavior in FGM plates for specific applications. The findings underscore the potential of FGM plates in achieving desired dynamic characteristics, thereby facilitating the development of lightweight, high-performance components with improved efficiency and reliability.

ABSTRACT. The microvasculature of the cervix plays a crucial role in the overall health and functioning of the female reproductive system. The concentration of RBCs within the microvasculature of the cervix holds significant diagnostic value. By analyzing RBC concentration, healthcare professionals can gain valuable insights into the physiological status of the cervix and detect potential abnormalities or pathologies. Cervical pathologies, such as cervical cancer, infections, and inflammation, can lead to alterations in RBC concentration and distribution. Monitoring RBC concentration can thus provide important clues for early detection and accurate diagnosis of these conditions. This paper highlights the importance of research focused on red blood cells in the microvasculature of the woman's cervix. The colposcopic images with polarized white light before and after the application of 5% acetic acid and Lugol's iodine were captured including cervicitis, vaginal bleeding, cervical ectropion, and cervical polyps. Based on the interaction of light with two-layer cervical mucosa tissue mode, a new image-processing algorithm is built to be linear to the concentration of red blood cells RBC in the ectocervix. Understanding the microvasculature of the cervix and its relationship with RBC concentration is of paramount importance for the effective diagnosis and management of cervical pathologies. The findings from these studies have the potential to improve early detection, prognosis, and treatment outcomes, ultimately enhancing women's reproductive health and well-being.

ABSTRACT. In this study, porous calcite samples were prepared through two processes: soaking to create pores and soaking to transform calcite with calcium hydroxide as the precursor. Calcium hydroxide powder was mixed with Polyvinyl Alcohol (PVA) particles ranging in size from 100μm - 300μm - acting as a foaming agent. The mixture was then blended at ratios of 7:3; 8:2; and 9:1 with a 3 wt% PVA solution before being cast into molds. The sample was soaked in distilled water at 80℃ to dissolve the PVA and leave pores in the structure of the material. The sample was then soaked in Na2CO3 at 60℃ for 7 days for phase transformation to calcite before being coated with chitosan. The phase composition of the sample was evaluated by X-ray diffraction (XRD). The mechanical strength of the sample was evaluated by Diametral Tensile Strength (DTS) testing. Porosity and density were evaluated according to ASTM C380-00 standards. The structure and morphology of the pores were evaluated using Scanning Electron Microscopy (SEM). For in-vitro experiments, the reaction activity of the sample before and after chitosan coating were conducted using simulated body fluid (SBF). The results showed that all samples were almost completely transformed into calcite with a total porosity of 70-80%, pore sizes ranging from 200-300μm, and DTS strength ranging from 25-60 MPa. In addition, the amount of PVA particles significantly affected the porosity and mechanical strength of the samples. Chitosan-coated samples had a total porosity of 65-70%, average pore sizes ranging from 100-200μm, and DTS strength ranging from 25-70 MPa. The properties of porous calcite materials are positively affected by the chitosan coating. This effect reinforces the properties of the material, and the combination becomes a composite material with many potential applications in biomedical materials engineering.

ABSTRACT. A real function defined on a nonempty and convex $\mathcal{D}\subset\R^n$, which remains quasiconvex under small linear perturbations, is \textit{s}-quasiconvex [Phu and An, Optimization, vol. 38, 1996]. The supremum of norms of these linear perturbations is said to be the stability index of the function. In this talk, we present the relationship between the stability index of a quasiconvex function on $\mathcal{D}$ and its stability index on line segments contained in $\mathcal{D}$. The closure of the set of linear perturbations above is proved. Additionally, approximation algorithms for computing the stability index of quasiconvex functions on compact and convex set $\mathcal{D}\subset\R^n$ are presented. Some examples of computing the stability index of some quasiconvex functions are shown.

ABSTRACT. In this paper, we are interested to study the inverse source problem for Poisson equation. Our models are concerned with two conditions: terminal data and integral nonlocal data. The problem is ill-posed in the sense of Hadamard. We show that the convergence of the source term when the parameter tends to zero. When the input data is noised, we use truncation method to regularize the problem in $L^2$ setting. We also give the error estimate in $L^p$ space.

ABSTRACT. Ensuring honesty and fairness in exams is an important and indispensable factor in the modern learning landscape, especially in national exams and particularly at Ho Chi Minh City University of Technology. To meet this need, we conducted experiments that demonstrated the simplicity, efficiency, and high accuracy of the Principal Component Analysis (PCA) algorithm in face recognition.

In this study, we explore the application of the PCA algorithm to develop software capable of verifying whether test takers are on the official list or not. By harnessing the power of PCA, we aim to provide a reliable and automated method for verifying the identity of test participants.

Our results show that PCA-based facial recognition is a good solution for participant verification, offering a simple and accurate approach to ensuring test integrity. This research contributes to improving the overall safety and fairness of academic testing, thereby promoting the principles of integrity and equity in education.

ABSTRACT. Cardiovascular diseases, especially those related to blockages, narrowing of blood vessels, or aneurysms, remain one of the leading causes of death among patients worldwide. Currently, artificial blood vessel grafting is one of the popular surgical methods to replace damaged native blood vessels. However, in the past several decades, artificial blood vessels still have several challenges, such as leakage, deformation, and blood vessel blockage. In this review, we employed the method of direct interaction between blood flow and the blood vessel wall, known as Fluid-Structure Interaction (FSI), to analyze the mechanical behavior of various artificial blood vessel models. The results obtained from this analysis will enable us to develop optimal designs that offer the highest efficiency in treating patients.

ABSTRACT. E-waste exposure to humans has been an issue for both developing countries. As technology has advanced, the production of waste has increased, and the toxicity of this e-waste has been giving workers adverse health effects. Liquid crystal monomer (LCM) is one of the toxic organic compounds within e-waste and has been researched extensively to figure out the degree to which it affects human health. Because this is a global concern, countries are currently trying to implement effective solutions.

We have identified previously written research papers related to the topic of e-waste and its effect on human health primarily by searching them in Pub Med, using specific search terms. Then, we filtered out the non-related papers by filtering out the preprints, retracted publications, and other animals (excluding humans). Moreover, we filtered out the papers by dividing them based on the categories of include, exclude, and review. This way, we were able to have a list of only the related papers for our review. After developing this list, we looked over the research papers in our list and reviewed what has been discovered about e-waste exposure and the harm and also about what should be done to further solve the issue.

The research papers proved the detrimental effects of e-waste on the human body. Specifically, they proved that LCM plays a major role in being the toxic component inside e-waste. Organizing sources based on search terms and filtering them through two different filtering methods allows relevant research papers to be gathered efficiently. Moreover, future studies would have to reveal further details of the e-waste management and LCM from the e-waste.

ABSTRACT. The electronic properties of the 2D hexagonal P63mc space group germanene are investigated using the density function theory method. The band structures and vibration properties of germanene, full hydrogenated germanene (FGeH), semi-hydrogenated germanene on top high buckling (SGeT1), semi-hydrogenated germanene on top low buckling (SGeT2) are analyzed. It was shown that the band structure of three models transits germanene from semimetal to metal and again back to semimetal, prooving that the less hydrogen adsorbed on germanene, the more likely it is metal. The phonon dispersion showed that germanene and full hydrogenated germanene (FGeH) were more stable than the other two models.

ABSTRACT. Beams are a common type of structure in mechanics that are used to model various objects, such as bridges, buildings, and aircrafts. Therefore, it is crucial to monitor the health and integrity of beams to prevent failures and accidents. However, most of the existing methods for health monitoring of beams rely on dynamic response, which is the change in the beam's vibration characteristics due to damage. A drawback of using dynamic response is that it involves a complicated process of data acquisition and analysis, and it also requires a large amount of memory resources. Hence, this paper proposes an alternative approach that uses static response with different positions to evaluate the damage level on beams. The results show that static deflection, can effectively identify both the location and severity of cuts that are present on beams. Therefore, static deflection can also serve as a useful indicator for health monitoring of this kind of structure.

ABSTRACT. By using the Molecular Dynamics simulation method, this study aims to show the influences of cooling rates on the solidifying temperature of the water inside a single-wall-carbon-nanotube (SWCNT) under different ambient pressures. We first created different systems with different tube diameters, then we cooled the systems from 300 K down to 200 K under different ambient pressures to observe the behavior of water. Our results showed that the more rapid cooling rate of the systems creates more disruptive and dramatic phase transitions that localize in specific ranges of temperature. Moreover, we also found that the lower pressures correlate to the more dramatic phase transitions of water molecules, regardless of the cooling rate. This study generally provides more insight into water behavior in the SWCNT with variations in ambient conditions.

ABSTRACT. Transiesta, an implementation of non-equilibrium electronic transport in the Siesta simulation program, was used to determine the current-voltage characteristic for phosphorene. The current-voltage feature comes as a curve on a two-dimensional plane, by plotting that characteristic of an electrical component, its electrical properties can be found. Therefore, we can correctly adjust the electrical component to use it efficiently. The two electrodes consisting of 36 phosphorus atoms each, have been relaxed with density-functional theory simulation and will be used in further electronic properties calculation. The scattering region has 5 lines of phosphorene, with the bias voltage increasing from 0V to 2V. The results of this simulation will help us look into the electrical properties of this phosphorene nano device.

ABSTRACT. Using ab initio calculations to study hydrogen adsorption on pristine germanene with a tetra structure. Performing convergent density functional theory calculations, the hydrogen atoms on Top 1, Top 2, Bridge, and Center positions were investigated. The Top 1 is found to be the most stable position. The adsorption of hydrogen atoms on tetra germanene changes the local structure of germanene. This is the first study on hydrogen adsorption onto tetra germanene, electrochemical properties were also investigated in this study.

ABSTRACT. Antibiotic residues in aquaculture wastewater have been knowns as contaminants because of long-term bioaccumulation that adversely affects the human/animal health and aquatic ecosystem. This study focuses on synthesis and studying the material properties of TiO2 nanotube arrays (TNAs) and TiO2 nanowires/nanotubes (TNWs/TNAs) by anodizing method using an aqueous NH4F/ethylene glycol solution for photocatalytic degradation of antibiotics (sulfamethazine, oxytetracycline, sulfamethoxazole and vancomycin). TNAs and TNWs/TNAs exhibited pure TiO2 anatase with (004)- and (101)-preferred orientations, and well-defined TNAs and TNWs/TNAs morphologies. The results show that TNWs/TNAs possesses higher photocatalytic activity than TNAs, primarily due to higher surface area of the former than the latter.

ABSTRACT. Vaccine refrigerators are essential guardians of public health. Their primary function is to create a controlled environment that safeguards the integrity of vaccines, ensuring they remain potent and safe for administration. Temperature control within these refrigerators is meticulous. Vaccines are delicate, and even minor temperature fluctuations can render them ineffective. Vaccine refrigerators employ advanced temperature regulation systems, maintaining a tight range of 2°C to 8°C (36°F to 46°F) for most vaccines. Some vaccines, like those for certain diseases or pandemics, require even stricter temperature control, often reaching ultra-low temperatures. Humidity management complements temperature control. Excessive humidity can introduce moisture into vaccine vials, potentially compromising sterility and potency. Moreover, humidity fluctuations can alter vaccine components, which may affect their effectiveness. Vaccine refrigerators work diligently to maintain humidity levels between 30% and 50%, creating a stable environment. Real-time monitoring and alert systems are vital features of these refrigerators. They continuously track temperature and humidity conditions and issue immediate alerts if there are deviations from the recommended range. This proactive approach allows healthcare providers to take swift action to prevent vaccine spoilage and ensure patient safety. Beyond real-time monitoring, vaccine refrigerators also record temperature and humidity data over time. This data serves as an invaluable resource for quality control, audits, and regulatory compliance. It enables healthcare professionals to review the entire storage history, ensuring vaccines' integrity from manufacturer to administration. In essence, vaccine refrigerators play an indispensable role in public health. They are the unsung heroes of vaccination programs, ensuring that vaccines retain their life-saving properties and contribute significantly to safeguarding global health.

ABSTRACT. The musculoskeletal system is an organ that generates force in the body, capable of rapidly changing shape and size in response to various stimuli. Different states of bodily activity affect the structural and functional changes in the musculoskeletal system. Lack of exercise can lead to muscle atrophy, while training with varying intensities can promote muscle growth. However, excessive training can result in muscle fatigue and stiffness, while low-intensity training may not achieve optimal results. Therefore, assessing the impact of exercises on the musculoskeletal system is crucial for its development. This report presents a computational modeling approach to investigate the response of the musculoskeletal system to various activities of the body: when lying down, during daily activities, and during resistance training. The results can be used to design appropriate exercises for musculoskeletal disorders.

ABSTRACT. Different sizes of the initial configuration are used to study the melting process of free-standing hexagonal Boron Nitride via molecular dynamics simulation. Initial configurations containing 10000 and 20000 atoms are heated up from 50 K to 7000 K via Tersoff potentials to have an entire picture about the structural evolution of free-standing hexagonal Boron Nitride upon heating. Various thermodynamic quantities are calculated to study the mechanism of melting process as well as the structural evolution, such as the total energy per atom, the heat capacity per atom, the radial distribution functions. The phase transition for both configuration sizes in this work exhibits the first order. The melting point of models depends on the initial size of the configuration in the range of this study.

ABSTRACT. The predictors of in-hospital mortality for patients admitted to the intensive care unit (ICU) have not yet been well characterized in spite of its importance in patients and risks stratification. Therefore, this study was designed to build a machine learning model and characterize different scenarios to predict outcome of in-ICU patients. A dataset of 1177 patients during their stay in the ICU was analyzed this in work. From this dataset, the demographics, laboratory test variables, vital sign, comorbidities and others information were utilized to predict mortality status of patient. In the first experiment, the importance factor of hematology, biochemistry, and urine laboratory were analyzed to evaluate the contribution of these variables to the predictors. In the second experiment, the imbalance-class challenge were handling by an oversampling technique. In the training state, 3 machine learning models were applied, which were Random Forest model (RF), XGBoost, support vector machine (SVM). The results presented the higher performance on model with oversampling technique, in which weighted F1-score were 84%, 83%, and 69% in RF, XGBoost and SVM respectively, while without oversampling technique the performance were 79%, 81% and 69% of weighted F1-score in RF, XGBoost, and SVM correspondingly. In conclusion, although there are many missing values, the results present a particular view of features importance, model performance of the mortality prediction in difference scenarios.

ABSTRACT. Stress is a prevalent psychological phenomenon experienced across all age groups. However, this adverse event is underrated and barely gets noticed when it reveals its harmful effects on the sufferer or when consulted by a specialist. Therefore, remote detection of stress plays an important role in improving living and working quality. This research aims to create a protocol to detect stress based on communication behavior, moreover, to discover early signs of stress before it enters a dangerous stage for the patient. The conversation data of 48 students obtained from mobile devices in the StudentLife dataset was analyzed in this work. Machine learning, statistical-based features were extracted from the conservation data to train 3 machine learning models: Random Forest, Support Vector Machine, and XGBoost. The results of XGBoost present the highest performance with an accuracy of 88% and a weighted F1-score of 77%. Moreover, an analysis of conservation behavior was conducted in this work. These results are beneficial for the remote protocol of stress controlling and provide remarkable insight for further research on smartphone data.

ABSTRACT. In the realm of healthcare, a plethora of data is available for extracting knowledge, leading to favorable outcomes. Specifically in medical field, Electronic Health Records (EHRs) encompass comprehensive information regarding a patient's health status, which are regarded as unstructured text documents utilized to preserve health by detecting, treating, and healing illnesses. Therefore, a appropriate analysis on EHRs could provide considerable information for diagnostic that surpasses the complex, time-consuming, and manually intensive analysis process of traditional methods. This research centers on the multi-classification of medical codes derived from EHRs. The main idea of this study is to use supervised learning techniques including XGBoost, Adaboost, Random Forest and Decision Tree Classifier models with the N-gram feature extraction method. The classification performance is assessed using the weighted F1 score and accuracy. Moreover, the Principal Component Analysis (PCA) was ultilized to reduce the dimension of the N-gram features. The highest performance achieved in model Random Forest with unigram feature extraction, which achived weighted F1 score of 0.94 and accuracy of 0.94 on the test dataset. This article covers Natural Language Processing (NLP) techniques and their evaluation methods. It also highlights the potential of NLP in healthcare, from data analysis to patient counseling.

ABSTRACT. The spinal column serves as the protective housing for the spinal cord, while also serving as the conduit through which peripheral nerves transmit control signals to the muscles. These neuromuscular pathways do not function in isolation but rather exhibit coordinated, rhythmic interactions to produce synergistic movements. Consequently, abnormalities in the spine can significantly impact an individual's mobility. The treatment of spinal-related conditions is often both time-intensive and costly. Conventional diagnostic tools such as X-rays, CT scans, and MRI scans primarily offer structural evaluations, lacking objective functional insights. Furthermore, X-ray imaging poses potential long-term health risks due to radiation exposure. This underscores the necessity for a safe alternative that reduces reliance on radiation-based diagnostic methods for structural assessments while providing objective functional evaluations. Functional assessments of the spine's impact can be effectively achieved through gait analysis, offering valuable insights for treatment. The aim of the present study was to apply formetric scanning methods to analysis of gait of patients with spine pathologies. The results show that characteristic anomalies, including reduced walking speed, increased ground contact time during various gait phases (stance phase, load response, pre-swing), decreased swing phase duration, and reduced single support time. Furthermore, the study demonstrated a diminishing range of motion in the hip, knee, pelvis, and spinal joints, along with an increased range of motion in dorsal and thoracic vertebral rotation. Specifically, during load response, hip flexion angles decreased, while knee angles exhibited variability among patients. This research hinglight the significance of gait analysis as a valuable tool for assessing functional spinal impacts, providing critical insights for designing effective treatment strategies.

ABSTRACT. In recent years, impedance pneumography (IP) has become a common approach for monitor respiratory rate. By injecting a stable small current into the subject’s ventral chest through a pair of electrodes, while measuring the voltage by the other pair of electrodes, the variations of thoracic electrical bioimpedance (TEB) can be monitored. The objective of this research is to create a circuit to monitor TEB and respiratory rate. A generated circuit involves: a current source and a data acquisition part. The results are TEB and respiratory rate. In conclusion, a circuit to monitor TEB and respitatory rate is generated successfully, which is an important platform for creating a wearable device to monitor respiratory rate constinuously.

ABSTRACT. Alzheimer's disease (AD) is one of the disorders that damages brain cells, leading to memory loss, cognitive difficulties, and forgetfulness, which collectively form Alzheimer's or commonly known as dementia. Currently, there is no effective therapy for AD; however, medication may slow the disease's progression. Therefore, early detection is essential to prevent AD from advancing to severe and life-threatening stages. Distinguishing healthy nerve cells from soft tissue in MRI images is a challenging task for doctors and radiologists, requiring considerable skill and time. Artificial intelligence methods can play a crucial role in early AD detection through MRI image analysis. In this study, our focus is on developing a Convolutional Neural Network (CNN) utilizing InceptionV3 as the base model to detect the four main stages using pre-processed MRI images. Our model achieved an area under the curve (AUC) of 0.8039 when using only InceptionV3 as the base model and 0.8258 when we pre-processed MRI images and added additional layers to our base model. The performance metrics indicate that our classification algorithm outperforms other current methods.

ABSTRACT. Currently, X-ray and CT scans are considered the gold standard for diagnosing spinal deformities. However, these radiation-based modalities can have adverse effects on a patient's health, particularly in adolescents, both during treatment and in the long-term. The primary objective of this study is to propose an alternative approach: the acquisition and reconstruction of spinal images using ultrasound techniques. This approach aims to reduce radiation exposure while maintaining the reliability of outcomes, comparable to those achieved through X-ray and CT scan techniques. This study was conducted using the curved probe of the Samsung HS40 ultrasound system. The scanning route was divided into several segments, with each segment approximately 6cm in length and captured in 2 seconds, at a sampling rate of 27 frames per second. Within each lumbar and thoracic region, the Transverse Process (TP) and Anterior Process (AP) of the vertebrae were possibly identified on 2D scans, which could significantly contribute to the scoliosis diagnosis procedure. Furthermore, after acquiring 2D slices of the spine, the coronal plane of the spine was reconstructed to estimate the Cobb angle. These results provide valuable reference points, similar to X-ray images, for guiding effective spinal deformity treatment. This alternative technique offers a promising avenue for a safer yet equally effective diagnosis and monitoring process.

ABSTRACT. For a better understanding of two-dimensional silicon carbide (2D-SiC) characteristics in fuel cell applications and semiconductor devices, the interaction between 2D-SiC and hydrogen was simulated using the ab initio method. Hexagonal prisms 2D-SiC have been created using the bond length obtained from molecular dynamics simulation 1. The number of hydrogen atoms adsorbed on top of Si and C atom, have been increased from two hydrogen atom to partially, and fully hydrogen coverage. The interaction energies between two hydrogens are shown to be attraction forces. Hydrogen is more likely to form bonds with neighboring hydrogen atoms, creating a cluster of adsorbed hydrogen. Three fully hydrogenated 2D-SiC models including table-like, chair-like, and boat-like were found stable, with the chair-like conformer being the most energetically favorable. Band structure and phonon vibration of those configurations was further analyzed, with the bandgap increasing in all the fully hydrogenated 2D-SiC conformer. This shows 2D-SiC can stable adsorbed hydro at a 1:1 ratio (4.7 wt%) and can be utilized in future tunable wide bandgap electronic devices.

ABSTRACT. The designs of target-drug delivery systems are attractively concerned due to their efficacy and safety. Fullerene is the first symmetrical carbon nanomaterial invented in the world. Due to the special properties of fullerene, it is an emergent topic in nanomaterials in recent years. Many experimental studies used this material to form the drug-carrier system and have shown a significant improvement in the pharmacokinetic properties of the active substance. Curcumin is a natural compound extracted from turmeric, with many pharmacological properties such as antiviral, antibacterial, and impact on cancer cells, etc. However, curcumin's pharmacological properties are hardly clinically demonstrated due to its water-solubility. A fullereo-curcuminoid derivative to HIV viruses and cancer cells was created, in which curcumin is out-bound to fullerene. HIV antiviral properties showed only moderate efficiency, and no anti-cancer effect was observed. Another disadvantage of the out-bound fullereo-curcuminoid derivative is that it is hard to control the number of curcumin-derivative molecules that bind out-surfaced fullerene, which is a critical problem we need to deal with since curcumin overdose causes side effects to the digestive system, skin, or headache. For the above reasons, we decided to conduct this research, focusing on the computational approach of in-bound fullereo-curcuminoid derivative systems for drug delivery, with adequate fullerene size to encapsulate curcumin molecules. This proposed model is promising not only to create a better antisolvent shield for the curcumin molecule throughout the delivery path to the target cells but also to manipulate the curcumin dose since the fullerene shield may increase the efficiency of curcumin carrying. This research uses the computational simulation method to investigate the epidermal growth factor (EGF) receptor binding and the physicochemical parameters of the curcumin molecule encapsulated in fullerene. The density functional theory (DFT) calculation is conducted to observe the electrical and energetic properties of the curcumin-fullerene encapsulation system. The obtained system is then docked with the target receptor. After that, the size-modified defected gap will becreated on the fullerene surface in the release process of the curcumin out of the fullerene. To interact with the target residues on the receptor will be observed by using MD simulation and their interaction stabilization.

ABSTRACT. Silicon carbide possesses a flat two-dimensional structure, making it a promising material for constructing twisted bilayer systems, which have various potential applications. This study utilized DFT calculations to analyze four models with different twist angles: 21.8°, 17.9°, 13.2°, and 5.1°. The objective was to assess how the electronic and phononic properties depend on the twist angle. The findings indicate that altering the twist angle can proportionally modify the band gap of bilayer SiC. However, the changes in band gaps are relatively small, with a mere 0.24 eV increase when the twist angle is adjusted from 5.1° to 21.8°. When the structure of each layer is fixed and the separation distance is reduced to 3.5 Å, 3.0 Å, 2.7 Å, and 2.5 Å at the four considered twist angles, the band gaps experience a significant decrease. Notably, this compression also causes the band to linearly decrease at a consistent rate, regardless of the twist angles. On the other hand, the value of the twist angle does not impact the phonon bands.

ABSTRACT. This paper aims to investigate the thermal properties of insulation materials made from natural resources depending on the increased temperature. Raw fibrous materials derived from natural resources or agricultural residues are currently being used as a prominent solution to significantly reduce thermal load and energy consumption in building and construction due to their benefits such as being lightweight, environmentally friendly, and biodegradable. In this research, the tested samples were manufactured from raw fibrous materials including coir, sugarcane bagasse, and luffa cylindrica fiber using the hot-pressing technique and the wet-forming method. The thermal conductivity coefficient was conducted at different levels of temperatures using the mean of the heat-flow-meter method according to the ISO 8301 standard. It is found that the λ-values of natural fiber-based insulation materials with the nominal density of 110-130 kg/m3 varied from 0.04 to 0.055 W/(m·K) demonstrating that they can be used as a good insulation material for building applications. The temperature dependency also showed that the higher temperature levels are always reflected in the higher values of thermal conductivity and an increase of 15-20% in thermal conduction regarding the increased temperature from 0 to 40 °C was recorded. Additionally, the thermal degradation regarding a temperature range of 30–800 °C was also practically investigated using thermal gravimetric analysis to observe the weight loss of the samples. Generally, the achieved results display superior potential for use as effective insulation materials in buildings.

ABSTRACT. Vibration analysis of continuous microbeams carrying a moving load is presented in the framework of Euler-Bernoulli beam theory and the modified couple stress theory (MCST) for the first time. The continuous beams consist of three spans with non uniform cross-section and simply supported ends. A finite element formulation is derived and used to construct the discretized equation of motion. Natural frequencies and dynamic response are determined with the aid of implicit Newmark method. The derived formulation is validated by comparing the obtained results with those published in the literature. The numerical investigation reveals the importance of the microstructural size effect on the vibration of the continuous microbeams, and incorporating the material length scale parameter in the formulation leads to an increase in the vibration frequencies, but a decrease of the dynamic response. The effects of the material length scale parameter and moving load velocity on the vibration behavior of the continuous microbeams are studied in detail and highlighted.

ABSTRACT. The Euler-Bernoulli beam theory is adopted in conjunction with modified couple stress theory (MCST) in this paper to formulate a beam element for size-dependent nonlinear bending analysis of a tapered cantilever beam subjected to end force/moment. The element is derived in the context of the co-rotational approach in which the internal force vector and the tangent stiffness matrix are firstly derived in an element attached coordinate system and then transferred to the global one by the transformation matrices. An incremental/iterative procedure is used in combination with the arc-length control method to compute the response of the microcantilever. The obtained results show that the formulated element is capable to model accurately the nonlinear response of the microcantilevers by just several elements. The obtained result reveals that the size effect plays an important role on the large deflection response, and the displacements of the microcantilever are overestimated by ignoring the influence of the micro-scale size effect. The effects of the material length scale parameter and the tapered ratio on the nonlinear behavior of the microbeam are studied in detail and discussed

ABSTRACT. Rainfall is a triggering factor leading to landslides, especially in regions where landslides often occur after long days of heavy rainfall. The previous studies only used a single rainfall map for landslide susceptibility assessment. However, this approach is unreasonable because rainfall is a time-variant data. To solve this problem, this paper uses time series rainfall data for landslide susceptibility assessment. The time series data of 1-day, 3-day, 5-day, and 7-day maximum precipitation corresponding to each year from 2016 to 2020 in the mountainous area of Quang Ngai province are collected. Furthermore, the time series of landslide sites from 2016 to 2020 along with other influencing factors are used to develop a landslide spatial prediction model based on the XGBoost method. The performance of the prediction model is assessed by statistical indexes as well as the ROC method. The obtained results indicate that all the cases using consecutive-day maximum rainfall data show excellent prediction ability. Of which, the model with 5-day maximum rainfall has the best performance. In addition, this paper also compares this approach to the previous approach using annual average rainfall. The testing result also indicates that the predictive accuracy of the cases with consecutive-day maximum rainfall is significantly higher than the case using annual average rainfall. Therefore, this paper recommends using the time series of consecutive-day maximum rainfall for landslide susceptibility assessments in this region as well as for other similar areas.

ABSTRACT. Proportional Topology Optimization (PTO) is a non-gradient topology optimization method which is simple to understand, easy to implement, and is also efficient and accurate at the same time. This method has just appeared but has achieved certain achievements in comparison with the others. In this paper, we use PTO to solve multi-material topology optimization problems. We consider the compliance problems satisfying the mass constraint and cost constraint. The elastic modulus is interpolated as Heaviside functions and cost is linear functions. The functions with scaling and translation coefficients are introduced to interpolate the elastic modulus and the cost properties for multiple materials with respect to the normalized density variables. Density filtering is used to remove checkerboard patterns. A threshold projection is applied for multi-material density in order to reduce the presence of intermediate ones. There are many interesting solutions in the comparison to the given results obtaind from the gradient-based topology optimization methods using SIMP (solid isotropic material with penalization) interpolation.

ABSTRACT. This paper investigates the buckling response of laminated composite plates based on the first-order shear deformation theory using the finite element method. The finite element model uses a 4-node quadrilateral element, with five degrees of freedom for each node. The first-order shear deformation theory is assumed for plate modelling in this study because of its straightforward formulation and computational efficiency. This theory takes the shear deformation of the cross-section into account as a linear function with respect to the thickness variable z. The study examines buckling analysis on various laminated composite plate problems with different shapes, loads and boundary conditions. All the problems are considered in the plane stress state. Additionally, parametric studies are performed to analyze the impact of the slenderness ratio and applied load on the deflection, stress, and critical buckling load for the laminated composite plates. The obtained results are compared with reliable results from other scientific research and show good agreement.

ABSTRACT. This paper investigate the dynamic behavior of wave energy buoys, which convert the energy from ocean waves into electrical energy. The motion of the buoy in response to the waves is critical for its ability to generate power. The modeling approach considers the buoy as a multi-component system, including a floating body, power take-off system, and control system, and takes into account various factors that influence its dynamic behavior, such as wave height, period, and direction. Experimental data are used to validate the model, and then used to investigate the buoy's sensitivity to different design parameters and operating conditions. Wave energy buoys are part of a class of wave energy converters that harness the energy from ocean waves to produce electricity. Their design and operation are challenging, requiring them to withstand the harsh ocean environment while efficiently capturing energy from the waves. Modeling the dynamics of the buoy is an essential step in this process. Ongoing research efforts in wave energy buoy modeling aim to improve their efficiency and reliability, making them a more viable source of renewable energy. Advanced control algorithms are being developed to optimize the buoy's performance by adjusting its motion in real-time while ensuring safety and stability.

ABSTRACT. This paper examines the nonlinear static behavior of plate structures subjected to bending load based on the first-order shear deformation theory (FSDT). FSDT is a simple plate theory that assumes only the first-order of the shear deformation, which makes the formulation easier to construct. Moreover, the use of a plate theory reduces the computational cost because a 2-dimensional model has fewer degrees of freedom than a 3-dimensional model to solve. The conventional finite element method (FEM) is employed in this study as the computational method. The nonlinear displacement-strain relation is considered in the study. First, a comparison is made with some existing data to show the accuracy and reliability of the method. Numerical examples are then presented of the influence of the load on the displacement. The results of the research can be applied to many different engineering applications related to plates subjected to bending loads.

ABSTRACT. Wind power towers are critical components of wind turbines that are subject to extreme loads, including cyclic loading, which can cause fatigue damage and affect their structural integrity. In this study, the National Renewable Energy Laboratory's (NREL) Mlife tool is applied to calculate the fatigue strength of wind power towers. MLife is a software tool developed using MATLAB to analyze outcomes from wind turbine tests, and dynamic simulations of aero-elastic behavior. It incorporates fatigue calculations such as damage rates and short-term damage-equivalent loads (DELs). These calculations are based on various factors including single time-series, lifetime DEL results based on the entire set of time-series data, accumulated lifetime damage, and time-to-failure. The results of the study can provide valuable insights into the fatigue behavior of wind power towers, which can be used to inform the design and maintenance of wind turbines.

ABSTRACT. In just the last two decades of the 20th century, many suspension bridges were successfully built in the world. Bridges with super-large span lengths and slender structures will be the main trend of research and development of bridge engineering in the coming decades. However, longer and thinner structures will face many difficulties, especially dynamics, earthquakes and aerodynamic behavior. It can be clearly seen that bridges with large span lengths will be very sensitive to aerodynamic influences and vibrations caused by wind. In recent years, a large number of suspension bridges have been and are being built in Vietnam. Vietnam is a country heavily influenced by wind and storms. Therefore, it is necessary to study the flutter instability of large span bridges. In this paper, the authors will analyze flexural and torsional vibrations with suspension bridge structure by analyzing fluid structure interaction with finite volume method through Ansys software.

ABSTRACT. In this paper, we investigate the damped pendulum model which is a classic in physics. Applying the Taylor expansion, we establish the qualitative properties of equilibrium points for this model.

ABSTRACT. We present the dynamics of fast two pulsed optical beam collisions in linear media with weak cubic loss that arises due to nondegenerate two-photon absorption. We investigate a perturbation method with small parameters to derive the general expressions for the collision-induced changes in the pulsed-beam’s shape and amplitude. Moreover, we design and characterize collision setups that lead to strong localized and nonlocalized intensity reduction effects. The predictions of the perturbation theory are in good agreement with results of numerical simulations with the perturbed linear propagation model, despite the strong collision-induced effects. Our results can be useful for multisequence optical communication links and for reshaping of pulsed optical beams.

ABSTRACT. We study the amplitude dynamics of flaton of the cubic-quintic nonlinear Schrӧdinger equation in the presence of randomness. We derive the SDE describing the amplitude dynamics of flaton and investigate the probability density function of the solitary wave’s amplitude. The methodology is mainly based on Ito’s calculus and Monte Carlo simulations. We extend this statistical behavior to other related solitary wave models.

ABSTRACT. In the context of infectious disease data analysis, the application of multifractal analysis, particularly Multifractal Detrended Fluctuation Analysis (MF-DFA), is explored, with a primary focus on understanding the COVID-19 pandemic. Daily case data from six countries is examined to unveil fractal behavior characterized by power-law relationships, offering valuable insights into the dynamics of disease transmission across various spatial and temporal scales. MF-DFA is introduced as a potent tool for analyzing nonstationary time series data, showcasing its ability to capture the intricacies inherent in natural processes. The study includes the computation of Local Hurst Exponents (Ht) at varying time scales, shedding light on local variations within the data. Additionally, the investigation of q-order Root Mean Square and q-order Hurst Exponents provides deeper insights into diverse aspects of data variability. This research underscores the multifractal nature of infectious disease data, emphasizing the importance of multifractal analysis in revealing nuanced patterns and correlations within complex time series data.

ABSTRACT. This article proposes a model of solidification of urea liquid droplets in urea prilling tower by using a two – phase Stefan problem with free moving boudary in spherical coordinates. The enthalpy method, which result is checked by comparing to the analytical solution of the problem, is used to simulate the prilling process and give conclusion about minimum heigh of towers so inside them droplets can be entirely solidified.

ABSTRACT. In this study we consider a stochastic SIR model in which the susceptible population is divided into subclasses. The long term behavior of the system is investigated under the influence of various system parameters. We construct a threshold for the convergence of the SIR model to the disease free case.

ABSTRACT. This talk proposes and develops a novel inexact gradient method (IGD) for minimizing C1-smooth functions with Lipschitzian gradients, i.e., for problems of C1,1 optimization. We show that the sequence of gradients generated by IGD converges to zero. The convergence of iterates to stationary points is guaranteed under the Kurdyka-Lojasiewicz (KL) property of the objective function with convergence rates depending on the KL exponent. The newly developed IGD is applied to designing two novel gradient-based methods of non- smooth convex optimization such as the inexact proximal point methods (GIPPM) and the inexact augmented Lagrangian method (GIALM) for convex programs with linear equality constraints. These two methods inherit global convergence properties from IGD and are confirmed by numerical experiments to have practical advantages over some well-known algorithms of nonsmooth convex optimization.

ABSTRACT. In this work, we investigate the problem of solving strongly monotone variational inequality problems over the solution set of the split feasibility problem with multiple output sets in real Hilbert spaces. The strong convergence of the proposed algorithm is proved without knowing any information of the Lipschitz and strongly monotone constants of the mapping. Inaddition, the implementation of the algorithm does not require the computation or estimation of the norms of the given bounded linear operators. Special cases are considered. Finally, a numerical experiment has been carried out to illustrate the proposed algorithm.

ABSTRACT. In this talk, we investigate the problem of solving strongly monotone variational inequality problems over the solution set of the split feasibility problem with multiple output sets in real Hilbert spaces. The strong convergence of the proposed algorithm is proved without knowing any information of the Lipschitz and strongly monotone constants of the mapping. Inaddition, the implementation of the algorithm does not require the computation or estimation of the norms of the given bounded linear operators. Special cases are considered. Finally, a numerical experiment has been carried out to illustrate the proposed algorithm.

ABSTRACT. This paper provides some second-order optimality conditions for efficient solutions to a nonsmooth multiobjective optimization problem subjected to equality and inequality constraints. We first propose a type of second-order Abadie constraint qualification to obtain a primal necessary conditions. A primal sufficient condition is also established. We next employ some alternative theorems to obtain the corresponding dual results. Applications to nonsmooth fractional programming problem are given. Some examples are provided to illustrate our results.

ABSTRACT. In this paper, we propose a new notion called radial directional derivative and derive its existence as well as main calculus rules. Then we employ them to investigate optimality conditions for a nonsmooth vector optimization problem subjected to an inclusion constraint in Banach spaces. With a directional (Hölder) metric subregularity assumption and a constraint qualification, necessary optimality conditions for both local weak and strict solutions are given in types of Karush–Kuhn-Tucker multiplier rules. The sufficient conditions are also established for local strict solutions whenever the decision space is finite-dimensional without any convexity assumption. Examples are provided to show advantages of the presented results over recent existing ones.

ABSTRACT. In this paper, we investigate an adjustable robust counterpart (ARC) of a two-stage uncertain linear problem. A non-adjustable robust form of (ARC) is derived and helps us give its solvable reformulated semidefinite programing (SDP) and evaluate its tractability. Under the local Farkas- Minkowski constraint qualification, optimality conditions and duality results are established. Some applications to robust counterpart (RC) and affinely adjustable robust counterpart (AARC) are obtained. These results are numerically illustrated by considering a practical problem with the support of some optimization packages.

ABSTRACT. We study global approximations in terms of certain types of variational convergence for a quasi-equilibrium problem and a generalized noncooperative game (Nash quasi-equilibrium problem). Set convergence of global solutions and numerical convergence of optimal value functions are established.

ABSTRACT. The automated defect detection system on industrial manufacturing lines in today's diverse world of consumer goods is a necessary requirement. Quality control is performed at various stages of a large-scale production process, including raw material and pre-production material inspections, in-process quality checks during production, and quality control of packaging, labeling before market releasing. In many industries, such as the garment industry, the quality of input materials significantly influences product quality, material utilization ratios, and ultimately the profitability of manufacturers. Fabric is one of the most critical input materials in the garment industry. However, during fabric production processes such as weaving, dyeing, and packaging, numerous factors can affect the quality of the raw fabric. Various fabric surface defects may occur, including yarn loss, yarn breakage, single yarn or area shrinkage, uneven dyeing, inconsistent color distribution, mold spots, and fabric thread breakage. These defects directly impact the final product and need to be eliminated during the classification process before entering production. Using manual labor to inspect each fabric roll with high accuracy becomes impractical in many cases due to several factors: experience, visual acuity, fabric roll speed, and psychological factors affecting operators' mental health from observing a monotonous surface for an extended period. All of these factors lead to the necessity of an error detection system on surfaces such as fabric. In this research, we introduce an approach to an optical system aimed at observing and detecting deviations in the fiber structure using images captured from a monochrome camera and a lighting system designed based on the actual structure of several types of fabrics used as research objects.

ABSTRACT. Seizures are a sudden and significant increase of electrical activities in your brain, affecting the daily routines of approximately 1% population all over the world. Modalities like functional and structural neuroimaging are common in diagnosing seizures. Although experiences with epileptic seizure diagnosis have indicated that the Electroencephalogram (EEG) method is most popular among doctors, it still has inherent disadvantages in which skilled physicians and a long time in observing its waves are typically unequivocal. Therefore, a considerable number of striking methods being proposed to support seizure diagnosis in healthcare facilities and research laboratories are ongoing and artificial intelligence is one of the most worthy to mention. Conventional methods in this area based on extracting features of EEG and using shallow networks like support vector machines, decision trees and so on have been bringing prospective results like cutting down observing time and providing reliable references for doctors in EEG waves-complicated situations. Besides, deep learning in the next years with outstanding results in a myriad of fields has gradually permeated into the EEG signals processing and has made an impression on tackling some drawbacks of its counterpart like avoiding overfitting, and automation in extracting features. In addition, more public datasets published on the Internet illustrate seizure diagnosis is not only captivating but also overwhelming. Last but not least, features in inputs of both methods also attract experts because they significantly affect providing accurate information about EEG signals which is the most important factor in diagnosing seizures accurately. In this paper, a well-rounded overview of seizure diagnosis methods using machine learning is provided with the aim of facilitating the following research and highlighting the most efficient methods in recent years.

ABSTRACT. Sintering or heat treatment is one of the common methods for fabrication of carbonate-substituted hydroxyapatite (CHAp) block for bone graft application. However, high temperatures make bulk CHAp reduce its bioactivity due to the decrease of carbonate content. The aim of this study was to evaluate the bioactivity of heat-treated carbonate-substituted hydroxyapatite in simulated body fluid solution (in vitro tests). According to in vitro tests, the bioactivity of heat-treated bulk CHAp depends on the carbonate content as well as the substitution type of carbonate group in apatite lattice.

ABSTRACT. Fast and green synthesis of ZIF-8 (zeolite imidazole framework) and its modifications are reported in two ways: TiO2@ZIF-8 and creating mesoporous. The samples were tested by some physic technology such as X-ray diffraction (XRD), Scanning electron microscopy (SEM), Infrared Spectroscopy meter (FTIR), and nitrogen adsorption measurements. The results showed clearly the mesoporous and TiO2@ZIF-8 found in polyhedral morphology. The adsoprtion of Tetracycline (TC) was thoroughly carried out by these hybrids ZIF-8, which showed high adsorption. For the first time, the number of molecules of TC adsorbed outer ZIF-8 surface area and inside its structure was calculated carefully. In general, the study performs a biocompatible synthesis of mesoporous and TiO2@ZIF-8, successfully applying for TC removal in solution.

ABSTRACT. Two stages were involved in the formation of carbonate apatite (CO3Ap). First, calcium hydroxide (CaOH)2 was used to fabricate porous calcite pellets, which were mixed with NaCl sacrificial granules of 150μm-250μm size in different ratios of 10:0, 9:1, 7:3, and 5:5. The mixture was homogenized, and an ethanol solution was added at a 6:4 ratio before casting into cylindrical shapes of 3x6 mm. Next, the calcite was immersed in a 1M Na2CO3 solution at 60℃ for three days and then in a 1M Na2HPO4 solution at 60℃ for four to seven days to transform into CO3Ap mineral. X-ray diffraction (XRD) and Fourier Transformed Infrared spectra (FTIR) showed that the precursor Ca(OH)2 had transformed into calcite and calcite had transformed into CO3Ap. Calcite samples had a total porosity of 45-70% with a DTS strength of 0.34-0.84 MPa. CO3Ap samples, which are derived from calcite soaked in a 1M Na2HPO4 solution, have a total porosity of 44-60% and a DTS strength of 1.26-3.42 MPa. The porous CO3Ap block was infused with ibuprofen to enhance its anti-inflammatory properties at the implant site. SEM images showed that ibuprofen crystals appeared in the CO3Ap porous block structure. In-vitro experiments were conducted to evaluate the activity of CO3Ap samples when immersed in simulated human body fluid (SBF). SEM images showed that all CO3Ap samples had apatite flakes and detected an apatite coating formed on the surface after soaking in SBF solution, indicating good activity. The activity of CO3Ap samples after loading ibuprofen in SBF solution for ten days was still maintained, confirming that ibuprofen loading did not affect the activity of CO3Ap. To improve the method of loading drugs into materials, two drug loading methods were used: soaking in a laboratory environment and in a vacuum environment. The drug loading efficiency and drug loading ability of material were then investigated using the UV-Vis analysis method. The study also examined the ability of porous CO3Ap blocks to release ibuprofen in Phosphate Buffered Saline (PBS) with pH = 7.4. The results indicated that the drug encapsulation efficiency and loading capacity were increased when using a vacuum environment. Therefore, the drug loading method was a crucial factor in the ability of the porous block to hold and release ibuprofen, which contributed to optimizing the process of manufacturing porous CO3Ap material carrying anti-inflammatory drugs for application as a bone replacement material.

ABSTRACT. Recently, in dentistry the infrared technique has been developed strongly to detect dental lesions, based on the fact that the optical properties of damaged tissue under infrared light are significantly different from those of sound dental tissue. Meanwhile, a number of research groups have paid attention to the application of deep learning strategies for classifying and analyzing infrared image data. The aim of the present study was to introduce the application of infrared technique and deep learning in detecting dental lesions. In this study, the optical systems with 850-nm LEDs which were applied for infrared imaging, were designed and used for diagnosing in vitro different types of early-stage lesions such as occlusal plaque, approximal plaque and white spot lesion. Two deep learning models (Unet and Mask R-CNN) were used for training and identifying the presence of lesions in 1367 near-infrared images. The results suggest the effectiveness of Unet and Mask R-CNN models in diagnosing dental lesions via infrared images.

ABSTRACT. Signal processing is essential for studies of bioelectricity, especially in the field of electroencephalogram (EEG). The objective of this research is to establish a process for analyzing data of EEG signals based on the main functional blocks, including artifactual signal removal, feature extraction, analysis of EEG signals using nonlinear space, data classification, and label noise evaluation of the results of classification. The research is carried out using three algorithms in order to extract the features, including Fast Fourier Transform (FFT), Short-time Fourier Transform (STFT), Wavelet and Power spectral analysis. The next step is to use the data after feature extraction as inputs transfering into functional blocks of automatic classification using algorithms of machine learning. In this work, Support Vector Machine (SVM) combined with Artificial Neural Network (ANN), Convolutional Neural Networks (CNN) along with Random Forest to classify fundamental states of the brain based on EEG signals. The authors will provide a process to analyze data of EEG based on functional blocks. Moreover, the research also proposes a method for validating label noise when the result is analyzed statistically. Overall, we have effectively filtered the signals and removed artifacts using different algorithms. Furthermore, the used algorithms have been proven to be able to achieve a high level of effectiveness in feature extraction and data classification. The proposed method also reached the accuracy at a definite degree in detecting label noise of the results.

ABSTRACT. Epilepsy is a chronic neurological disorder stemming from irregularities in the brain that trigger simultaneous activation of clusters of nerve cells, leading to abrupt and uncontrolled electrical discharges. The World Health Organization (WHO) reports an annual global estimate of 5 million individuals diagnosed with epilepsy, underscoring the urgency for innovative early detection techniques for epileptic seizures. The medical community is increasingly drawn to this pursuit, with a keen interest in harnessing Machine Learning approaches to discern seizure patterns from Electroencephalogram (EEG) signals. On an international scale, numerous cutting-edge investigations have been undertaken to refine the efficiency and precision of seizure identification processes. These endeavors contribute significantly to the accurate diagnosis and effective treatment of patients. Nonetheless, challenges persist both globally and within Vietnam. Principal among these challenges is the diverse nature of EEG data across various patients, necessitating the formulation of robust and adaptable methodologies for the extraction and classification of EEG signal features within the context of heterogeneous data environments. Additionally, any Machine Learning model developed must demonstrate mobility and the capacity for automation to ensure practical real-world implementation. This research paper presents a comprehensive methodology encompassing EEG signal preprocessing, leveraging the Short-Time Fourier Transform (STFT) for extracting pertinent spectral attributes, and training a Support Vector Machine (SVM) classifier to differentiate EEG segments with and without epileptic seizures. Evaluation of the proposed approach on a dataset showcases a classification accuracy of 80%, accentuating its efficacy in identifying epileptic seizures. The prospects of this research extend to its potential application within EEG signal processing systems in Vietnamese hospitals and clinics. This application could enable the early identification and prompt alerting of impending seizures, facilitating timely intervention and minimizing risks to patients' well-being.

ABSTRACT. During and after engaging in physical exercise, many people experience a common condition known as muscle tiredness. It is characterized by a loss in muscular force output and may have a severe influence on an individual's capacity to conduct day-to-day tasks as well as their sports performance. Electromyography (EMG) signals, which are electrical impulses created by muscles during contraction, may be used to evaluate muscular exhaustion. These signals are generated by the muscles. Researchers are able to design techniques to alleviate the detrimental consequences of muscle tiredness by evaluating changes in EMG signals that occur during physical exercise. This allows the researchers to acquire insights into the processes that cause muscle fatigue. The purpose of this study is to assess the present level of research on the use of EMG signals to measure muscle fatigue and to recommend areas for further research and development.In this study, the classification performance of Support Vector Machines (SVM) was evaluated combined with feature selection techniques: Principal Component Analysis (PCA).  In total 10  subjects were collected with 3 exercises and 3 repetitions. A total of 14 common features in muscle fatigue classification were extracted and selected by the above algorithms.The results of our study show that SVM classification F1-score specifically 0.06339.During and after engaging in physical exercise, many people experience a common condition known as muscle tiredness. It is characterized by a loss in muscular force output and may have a severe influence on an individual's capacity to conduct day-to-day tasks as well as their sports performance. Electromyography (EMG) signals, which are electrical impulses created by muscles during contraction, may be used to evaluate muscular exhaustion. These signals are generated by the muscles. Researchers are able to design techniques to alleviate the detrimental consequences of muscle tiredness by evaluating changes in EMG signals that occur during physical exercise. This allows the researchers to acquire insights into the processes that cause muscle fatigue. The purpose of this study is to assess the present level of research on the use of EMG signals to measure muscle fatigue and to recommend areas for further research and development.In this study, the classification performance of Support Vector Machines (SVM) was evaluated combined with feature selection techniques: Principal Component Analysis (PCA).  In total 10  subjects were collected with 3 exercises and 3 repetitions. A total of 14 common features in muscle fatigue classification were extracted and selected by the above algorithms.The results of our study show that SVM classification F1-score specifically 0.06339.

ABSTRACT. Over the period of the last few decades, there has been a rise in study into Digital Image Processing, especially in the sports industry. Specifically, this method has been used in table tennis to collect data from practice sessions or matches, in order to evaluate the facts of the match or the performance of the athletes. This evaluation helps athletes improve and understand their performance by identifying their strengths and weaknesses. However, due to high requirements of spaces and costs for installation, Digital Image Processing is mostly used for professionals area. As a consequence, the objective of this article is to represent a Digital Image Processing application for table tennis training with ball launcher. This application is able to keep track of each and every ball that is fired from the launcher, with reasonable accuracy, and determine where the ball will bounce when it is returned to the table by player's hit. After each practice session, the player's results will be shown and stored for long term progress tracking. The ratio of balls fired to balls returned and dropped on the other table side will determine the player's effectiveness. This application makes use of a variety of methodologies, some of which are background subtraction, image masking, contour detection, and vector distance calculating. However, even if the application generates practice results, such outcomes can still be influenced by elements like the lighting in the room, the frame rate of the video, and other external factors. Thus, this application's algorithm must be improved to achieve the best results in as many diverse practicing environments as possible while improving its tracking speed and accuracy in real world scenarios.

ABSTRACT. The demand for advanced materials that can boost the effectiveness of structural applications has increased due to the additive manufacturing industry's incredible growth in recent decades. Due to their specific ability to provide energy absorption, lightweight properties, and multifunctionality, triply periodic minimum surface (TPMS) structures have become an advanced strategy. The unit cell geometry of TPMS isolates it from other porous structures like honeycombs. It significantly impacts a lattice structure's mechanical characteristics and energy absorption capability. The smooth and repeating unit cells that TPMS structures support the creation of high pore interconnectivity and surface-to-volume ratio. However, extensive analytical or numerical procedures could be more practical due to the complexity of TPMS manufacture. We use a technique called "Homogenization" to examine TPMS structures to deal with this problem. By substituting a similar homogenous medium for the composite material, homogenization produces macroscopic behavior consistent with the material's typical behavior. In particular, the Gyroid, Schwarz, and Diamond TPMS structure types are the three that this research concentrates on. We examine their mechanical characteristics and consider how they might be used in structural engineering. The analysis shows that the homogenization method provides a more efficient mathematical analysis with a manageable error rate. Furthermore, TPMS structures have great potential for actual use in various industries. TPMS structures can improve the effectiveness and performance of structural applications by utilizing their energy absorption properties, lightweight design, and multi-functionality. This study offers insightful information on the mechanical behavior and potential applications of TPMS structures, which aids in their comprehension and use in engineering design.

ABSTRACT. This article presents a proposition for employing a data generator technique to build a dataset for training an Artificial Neural Network (ANN) in solving the inverse kinematics (IK) problem of Selective Compliance Assembly Robot Arms (SCARA). The inverse kinematics problem involves determining the joint angles required for a SCARA robot to reach a desired end-effector position. To generate the dataset, we employ a random sampling approach. By randomly sampling the joint angles and corresponding end-effector positions within a feasible range, we obtain a diverse dataset of SCARA configurations. This dataset then trains an ANN, specifically a multi-layer perceptron architecture with multiple hidden layers and dropout regularization, to predict the joint angles for a given end-effector position. Using a data generation technique allows us to capture various SCARA configurations and their corresponding solutions. This enhances the generalization ability of the ANN and improves its accuracy in predicting the joint angles. Training the ANN on this generated dataset enables it to learn the complex relationships between the input end-effector positions and the output joint angles. Furthermore, we incorporate Physics-Informed Neural Networks (PINNs) into the ANN-based solution to account for the physical constraints of the SCARA robot. PINNs utilize a custom loss function that combines data loss, position loss, and limit loss to ensure accurate and physically plausible predictions. This integration of physical constraints enhances the reliability and realism of the predicted joint angles. Combining data generation techniques and integrating physical constraints through PINNs provides an efficient and accurate approach to solving the inverse kinematics problem of SCARA robots. This approach has significant applications in industries such as manufacturing and robotics, where precise control of SCARA robots is essential.

ABSTRACT. Bridges, an essential engineering component, play a vital role in carrying and distributing dynamic vehicle loads. The utilization of beam structures in bridge models significantly impacts the performance and safety of structures exposed to dynamic loads. In this article, the dynamic analysis of beam structures subjected to moving loads using ANSYS Workbench software is conducted. The Euler – Bernoulli beam theory is assumed for the beam model, while the moving loads are represented as concentrated loads that traverse along the beam under loading conditions. Both static and dynamic simulations of the beam using ANSYS Workbench software are performed to evaluate the structural response under different loading scenarios. Various loading velocities are considered in both subcritical and supercritical regimes to assess the effects of different speeds on the beam's behavior. The results from these simulations provide valuable insights into the beam's deflection, which are compared against previously published simplified test conditions and produce fairly good results. The influence of the damping factor is also taken into consideration when evaluating the gradual diminution of the beam's displacement after the load moves on and away from the beam. Damping plays a crucial role in real-world structural behavior as it accounts for energy dissipation and determines the gradual attenuation of the beam's displacement after the load is removed. The research presented in this paper contributes to an academic approach to solving complex problems related to beam behavior under dynamic loads. Based on the beam's displacement, it can help predict the possibility of damage that may occur to the structure. The findings of this research can be applied to improve the design and performance of structures subjected to dynamic loads. This can result in safer, more resilient structures that ensure the safety of vehicles passing through them.

ABSTRACT. Building information modeling-BIM during the building construction and lifecycle monitoring brings many benefits to healthcare facilities. Building information modeling has been taken place around the world in general and Vietnam in particular for many years. Firstly, architects, contractors, civil engineers, biomedical engineering engineers, engineering mechanical engineers can work together on attractiveness, aesthetics, constructability and other aspects on BIM, in order to meet the routine needs of patients, doctors and nurses in healthcare facilities. Secondly, the effectiveness of BIM is specifically shown during the design process and virtual simulation of the building before and after construction, or during the operation and improvement process of healthcare facilities for the period of 30 years, 60 years or 90 years. Finally, studying the vibration analysis and quality control of BIM for the hospitals in Ho Chi Minh City is presented herein.

ABSTRACT. The study aims to introduce a beam finite element for analyzing the static and dynamic behavior of micro beams, with a specific focus on size-dependent effects. The proposed model is based on the Modified Couple Stress Theory (MCST) and follows an Euler-Bernoulli beam formulation. The finite element method is employed using the Galerkin technique to solve the governing equations of motion and boundary conditions for micro beam problems. The key aspect of the new element is its ability to account for size effects by incorporating a length scale parameter in the element matrices. When the parameter is set to zero, the element reduces to the Classical Euler-Bernoulli beam element. To validate the accuracy of the model, static and free vibration analyses are conducted under various boundary conditions. The comparison demonstrates good agreement, confirming the reliability of the proposed model. Additionally, a case study is presented, focusing on a micro-structure composed of interconnected micro-beams, emphasizing the significance of considering size-dependent behavior in the design and analysis of micro-scale structures.

ABSTRACT. Hyperelastic materials like rubber are primarily common in real life as well as industry applications, and it is still an active research area. Naturally, the characteristic of hyperelastic material will be expressed when it undergoes large deformation, so the geometrical nonlinear effect should be considered. The geometrical nonlinearity eliminates the small strain assumption, which makes the displacement-strain relationship become linear. To analyze the behavior of hyperelastic material, many models were proposed. The Neo-Hookean model is imposed in this study because of its simplicity. The model shows the nonlinear relationship when the deformation becomes large due to the nonlinear displacement-strain relation. This constitutive relation also gives a good correlation with experimental data. This study performs a meshless method, namely the radial point interpolation method (RPIM), to analyze the nonlinear behavior of Neo-Hookean hyperelastic material under a finite deformation state in three-dimensional space. The standard Newton-Raphson technique is applied for obtaining the nonlinear solutions. Different from mesh-based approaches, meshless shows its advantage with large deformation problems by its mesh-independence. In this study, the numerical results of three-dimensional problems that undergo large deformation will be calculated, and validated with solutions derived from previous studies.

ABSTRACT. Metal injection molding (MIM) is a process technology that combines the advantages of plastic injection molding and powder metallurgy technology, which is very suitable for manufacturing metal parts with high strengh, high precision and complex geometry. The manufacturing process covers a wide range of technologies, including mold design, mixing of powder and binder, injection molding, degreasing, sintering and secondary processing in the post-process. The success of mold design will determine the success of products development, which will directly affects the efficiency and quality of MIM production. However, the conventional 2D cooling channels may not be effected with the problem of non-uniform temperature distribution for multi-cavity molds, and may cause over heated, shaping defects of the product. In this study, we develope an integrated technology for the MIM product shaping and manufacturing, including MIM mold development technology for mold flow analysis, 3D conformal cooling channel design, metal additive manufacturing and CNC processing. Finally, through the optimized design of the 3D conformal mold cooling system, the cooling efficiency of the MIM product and the uniformity of the mold temperature are improved, which not only successfully improves the product quality but also increase the working cycle of the molding process.

ABSTRACT. In tool machines, ball screws are used to improve positioning accuracy as ill as to elongate their service life. Ball screw is the most widely used mechanism to convert rotational motion to linear motion in high-speed machine tools because of its high rigidity, efficiency, and accuracy. Usually, ball screws will be provided by the manufacturer with a preload level to achieve a certain accuracy. In practical establishment of a machine tool, the ball screw is usually preloaded via using oversize balls for a single nut, an offset pitch single nut or a double nut to eliminate backlash of machine movement and increase the stiffness of the drive system for precision motion control. However, after a long time of operation, this presetting preload will gradually degenerate because rolling and sliding friction wear out the balls, nut or screw of the drive system. The preload loss then may lead to a loir stiffness, strong vibration, inaccurate positioning and even fail of the machine tool. Thereby, increasing the stiffness of the ball screw lead to increase in its vibration frequency. This paper verifies the change in ball screw feed drive system stiffness according to changes in preload levels based on empirical formulas. This paper presents a method to increase the preload of the ball screw system by using 2 ball nuts and inserting a helical spring. In particular, the paper combines Hertzian contact theory and spring theory to create preload. In addition, the paper uses empirical formulas from previous studies to verify the change in ball screw feed drive system stiffness according to the change in preload through the axial natural frequency. The results are compared and evaluated with experiments from which conclusions and development directions are drawn. Hopefully, the results in this paper can be applied in practice in the field of ball screw feed drive system design and calculation.

ABSTRACT. The forward kinematic problem has the main task of determining the position and direction of the work joint relative to a fixed global coordinate system, which is a function of the matching variables. The theory method used to solve this problem is the Denavit-Hartenberg (D-H) matrix method. The objective of this survey is modeling and calculating the translation motion of the robotic arm with 6 degrees of freedom - DOF (Denso VS-6556 from Denso Wave Incorporated). First, the robotic arm is modeling into stages, link bars, motion joints and fixed dimensions for determining the (D-H) kinematic parameters. From the (D-H) kinematic parameters, we can determine the Denavit-Hartenberg homogeneous coordinate transformation matrix for each stage, especially the final stage. By using calculation tool (Matlab) and simulation tool (Solidworks) software, the robot arm's stages include position and direction relative to a fixed global coordinate system could be found. Thereby, plotting the displacement trajectory of robot stages based on time-varying joint variables. The authors compare the results from the data obtained between the two tools to evaluate the error and correctness of the kinematic survey method from the forward kinematics problem (D-H). In addition, from this model, there will be a survey of possible errors in the operating orbit through velocity and acceleration. The results are presented by comparing the results of two computational and simulation tools, opening a discussion about the influence of velocity and acceleration on the trajectory of an industrial manipulator. This can be a premise to solve further problems in future work.

ABSTRACT. In this work, an improved non-sensitivity structural topology optimization method incorporates virtual elements with unstructured polygonal meshes. Specifically, the recently developed gradient-free proportional topology optimization (PTO) is employed with a material distribution formula suitable for stress-constrained problems. We integrate a linear virtual element method (VEM) module into the PTO algorithm to improve performance. The linear VEM has a desirable characteristic: it does not require numerical integration and, more importantly, is less sensitive to small-edge elements, thus significantly reducing computational time. Additionally, we utilize virtual elements with unstructured polygonal meshes inside PTO to suppress one-node connection instabilities. We demonstrate the performance of our developed method through several benchmark examples of stress-constrained.

ABSTRACT. This talk presents a new version of Newton's method named Backtracking New Q-Newton's method, which has both good theoretical guarantees and experimental performance. We first introduce the main theoretical results for the method, its algorithm, in the general setting. To illustrate the usefulness, we then present experimental results (in both deterministic and stochastic) root finding. We then explain some interesting features of the experimental results and pose some questions.

ABSTRACT. In this talk, we present some new explicit criteria for exponential stability of positive continuous-time difference systems with time-varying delays. The obtained criteria include some existing results as our particular cases. Some examples are given to illustrate the obtained results. For more details, please refer to [Int. J. Control, 2023, Vol. 96, No. 6, 1650-1660].

ABSTRACT. In this paper, we study H¨older type estimates of approximate solutions to parametric vector equilibrium problems. Our results are improvements of the ones in the literature. An application of the main results to a weak traffic equilibrium problem is also presented.

ABSTRACT. In this talk, we are concerned with set-valued equilibrium problems under perturbations of both objective functions and constraints. By using a scalarization method, we obtain sufficient conditions for the Holder continuity of solution maps to parametric set-valued equilibrium problems.

ABSTRACT. Nonlinear time-varying differential systems with infinite delay represent a complex and challenging class of dynamical systems with broad applications in various scientific and engineering disciplines. This paper investigates the stability analysis of such systems and presents explicit criteria for exponential stability. Furthermore, robust stability bounds are derived to account for the presence of nonlinear time-varying perturbations, which are often encountered in real-world scenarios. The obtained results are illustrated through three compelling examples, showcasing their applicability and effectiveness. The novelty of this work lies in its contribution to the understanding of nonlinear time-varying systems with infinite delay, providing valuable insights for researchers and practitioners in the field. These findings open up new avenues for the control and optimization of dynamic processes characterized by infinite delay and nonlinearities, promising potential advancements in various technological applications.

ABSTRACT. In statistics and machine learning, classification is a topic of great interest to scientists due to its diverse applications. Although many classification methods have been developed, the Bayesian approach still has numerous advantages. This research focuses on improving the classification problem using the Bayesian method while enhancing prior probabilities and probability density functions (PDFs). With PDFs, we integrate Vine Copula and PDF components as kernel density functions to address dependencies among variables. For prior probabilities, the research employs fuzzy clustering techniques for determination. By combining these two enhancements, this researcher developed a further classification algorithm. The proposed algorithm is detailed in its implementation steps and illustrated by numerical examples. It is also applied in image classification for medical diagnosis. Numerical examples and applications demonstrate the effectiveness of the proposed classification algorithm compared to other common algorithms, including statistical and machine learning methods.

ABSTRACT. Electrical impedance spectroscopy is a non-invasive technique in the field of breast tissue classification owing to its cost-effectiveness, radiation-free, and early detection. The present paper uses a public data set of 106 cases representing 6 classes of excised breast tissue along with 8 electrical impedance spectral features. In this work, we use regression analysis with the multinomial logistic model to identify the effects of the electrical impedance spectrum features to the breast tissue types. The study also contributes classification techniques using two machine learning-based methods including support vector machine, and random forest. The classification is examined in two scenarios: the case with data with only two types of breast tissue and the case with six types of breast tissue. In addition, rather than concentrating solely on accuracy as in public papers, we also provide some other metrics to assess the classification models, such as AUC and F1 score. The study confirms that the random forest classifier exhibits more competitive results in comparison to such findings in other publications, as demonstrated by approximately 94% accuracy, 82% F1 score, and 88% AUC in a case of two tissue types categorization. In a second case where differentiating between the six various forms of breast tissue is significantly trickier, the random forest is still the best option for a classifier.

ABSTRACT. Multiple logistic regression is a widely used classification model, especially in clinical analysis. It makes use of probabilistic estimations, which aid in understanding the relationship between the dependent variable and one or more independent factors. As one of the most prevalent illnesses in the world, diabetes, if caught early enough, maybe stopped from progressing and additional problems may be avoided. In this research, we develop a prediction model, which uses specific diagnostic metrics from the dataset, predicts if a patient has diabetes, and uses other machine learning techniques to make a comparison of the performance and accuracy. Multiple logistic regression is the main algorithm used in the research and carried out using RStudio. The experiment mainly uses one dataset, which is the PIMA Indians Diabetes dataset from the National Institute of Diabetes and Digestive and Kidney Diseases. The accuracy of multiple logistic regression can be seen around 75 percent. While comparing other metrics with other machine learning methods, multiple logistic regression has shown to be one of the efficient algorithms in building prediction models. This study also suggests conducting data preprocessing rather than building a complex model, which results in higher computational costs.

ABSTRACT. Criterion-referenced tests (CRTs) are widely regarded as effective tools for assessing student learning and evaluating the teaching and training process.

This research focuses on techniques for preparing CRTs, including methods for simulating the difficulty of multiple-choice tests and measuring student ability through the results of some specific multiple-choice exams. We also present the analysis of fit for a multiple choice test by using the Rasch model.

ABSTRACT. A criterion referenced test is considered an extremely effective tool for assessing the learning process and therefore it is useful for evaluating the training and teaching process. The main aim of this research is to discuss about the technique of making a criterion referenced test. Some methods are used to simulate the difficulty of a multiple choice test, enabling us to measure the ability of a student through the result of a specific multiple choice exam. Finally, we conduct the analysis of fit for a multiple choice test via the Rasch model.

ABSTRACT. In this talk, we shall discuss about the long time dynamics for a double free boundary system with nonlocal diffusions, which models the infectious diseases transmitted via digestive system such as fecal-oral diseases, cholera, hand-foot and mouth, etc... We start by proving the existence and uniqueness of the Cauchy problem, which is not a trivial step due to presence of couple nonlocal dispersals and new types of nonlinear reaction terms. Next, we provide simple conditions on comparing the basic reproduction numbers R0 and R∗ with 1 to characterize the global dynamics, as t→∞. We further obtain the sharp criteria for the spreading and vanishing in term of the initial data. This is also called the vanishing-spreading phenomena. The couple dispersals yield significant obstacle that we cannot employ the approach of Zhao, Zhang, Li, Du [JDE-2020] and Du-Ni [Nonlinearity-2020]. To overcome this, we must prove the existence and the variational formula for the principal eigenvalue of a linear system with nonlocal dispersals, then use it to obtain the right limits as the dispersal rates and domain tend to zero or infinity. The maximum principle and sliding method for the nonlocal operator are ingeniously employed to achieve the desired results.

ABSTRACT. One of the fundamental questions in population dynamics concerns the criterion for the persistence or extinction of a biological species subjected to their habitat changes. In this communication, to understand more clearly the impact of climate change on the global dynamics of mosquitoes proposed in Bao (JMB 76:841-875, 2018), we consider a reaction-diffusion free boundary model with conditional dispersal in a heterogeneous environment. Our main interest is to study long-time dynamics of solutions assuming that the environment is globally unfavorable determined by a spectral condition at infinity. The mathematical models to describe the dynamics of a population facing climate change have arisen many challenges in science and application and our result makes a theoretically substantial contribution besides the previous works (Bao in JMB 76:841-875, 2018; Monobe in JDE 261:6144-6177, 2016; Shen in JMB 84:30-42, 2022; Shen in JDE 269:6236-6268, 2020; Vo in JDE 259:4947-4988, 2015) for the study of the impact of the climate change with the conditional dispersal and free boundary.

ABSTRACT. In this problem, we consider a Sobolev equation with the Atangana-Baleanu-Caputo fractional derivative. We study the existence of mild solutions to these equations. Using the embedding theorem, we provide an estimate of the error between the exact solution, and the normalized solution (by means of Fourier series truncation), with observed data in L^r(Ω)

ABSTRACT. In this report, we study the initial value problem for the fractional diffusion equations with the Caputo derivative and a source function satisfying some given assumptions. We establish the existence and uniqueness of mild solutions corresponding to two different initial data assumptions. We derive global results of a unique mild solution with small initial data using some Sobolev/SobolevOrlicz embeddings, a weighted Banach space, and the fixed point theorem. We also discuss the stability of the fractional derivative order for the time under some assumptions on the input data. In the absence of any smallness assumptions, the Cauchy iteration method demonstrates that the mild solution blows up at a finite time or exists globally in time. Finally, we consider some illustrated examples to test the results obtained in theory.

ABSTRACT. This talk is about a Cauchy problem for a non-focal Fisher-KPP equation. We demonstrate in our study that as long as the habitat limit of the considered population (with the density described by the solution) is large enough relative to the growth rate, there is always a unique global solution to the problem regardless of the size of the non-negative initial data. The idea can be outlined as follows. First, we prove the local existence and uniqueness of the mild solution. Second, we improve the temporal regularity of the solutions and show that the non-negativity of the initial data is preserved for this solution. Having proved these preliminary steps, we derive an energy estimate by which we can control the solution for all time.

ABSTRACT. The $\overline{\partial}$ operator is an important concept in complex analysis. This note gives a summary of $L^2$ techniques for the $\overline{\partial}$ equation from some works of B.Y. Chen. This is a report of known material.

ABSTRACT. We consider the model of van der Waals fluid flows in a nozzle with discontinuous cross-sectional area. First, the hyperbolicity is investigated. The model is elliptic-hyperbolic and the pressure function admits two inflection points. Second, we investigate the nonclassical shock, which violate the Liu entropy condition and satisfy the entropy dissipation inequality

ABSTRACT. This research focuses on exploring the existence of a nontrivial weak solution to a class of elliptic equations controlled by double phase operators with variable exponents on a bounded Lipschitz domain. The difficulty in our problem arises from the lack of compactness due to the presence of a nonlinearity exhibiting critical growths. We address it by borrowing the critical embedding proved in [Ho-Winkert, 2023] and the concentration-compactness principle stated in [Ha-Ho, 2023]. Aside from the existence, we also get the behavior of solution via variational method.

ABSTRACT. The research interest in establishing separability within semigroups has been inspired by Professor M.M. Leshokhin's works, as well as the contributions of four authors: Craig Miller, Gerard O'Reilly, Martyn Quick, and Nik Ruskus, who presented their findings in [1] and [10]. Our research builds upon this foundation and aims to delve deeper into the concept of P-separability within semigroups, where P represents various predicates. The investigation is not confined to finite semigroups but encompasses a broader spectrum. Specifically, we focus on the challenge of identifying a class K of semigroups (including infinite ones) such that a given class S of semigroups can be rendered P-separable through mappings, such as homomorphisms, characters, generalized characters, and more, from the set S into the set K. The problem of minimization of P- separability is also considered.

ABSTRACT. The concepts of width of a compact set $S$ at a point $p$, denoted by $W(p)$ in $\mathbb{R}^n$, are introduced. The continuity of the function $W(p)$ is proven. Furthermore, the width of $S$ with respect to a fixed subset $G$ of the set are defined.

In this paper, we consider the regularity of the continuity at a point $p$ that is in the interior of $S$ and at the point $p$ that is near the boundary of a convex set $S$.

ABSTRACT. We study the ill-posed backward problem for a contaminated nonlinear predator-prey system whose velocities of migration depend on the total average populations in the considered space domain. We propose a new regularized problem using quasi-reversibility method. Moreover, under some mild assumptions on the true solution, we give useful and rigorous error estimates and convergence rates and give conclusion about the stability of the regularized solution.