7264 research outputs found
Sort by
DESIGN AND OPTIMIZATION OF A REGENERATIVE BRAKING SYSTEM
No full textThe research was conducted to explore a unique modern method of recuperation called the regenerative braking system. The concept behind that is to generate voltage from the source of heat. The device that was used in our case is a Thermoelectric generator(TEG). This device converts the temperature difference into the potential difference.
As the braking is performed, there is an enormous amount of energy that is just dissipated in the rotor, and released to the environment. The study explores the way to implement the TEG into the braking system, find the best possible assembling approach, and discover whether the output parameters are going to be sufficient for the implementation of helpful systems.
In order to achieve the aims that were set, simulations, as well as hand calculations were applied. Results were obtained by using different softwares such as ANSYS, Solidworks, MATLAB and Excel. From the presented data, the analysis of the selected design was executed. Figuring out the applications that can be theoretically implemented with the help of regenerative braking systems, as well as the conditions in which they are the most effectiv
OPTIMAL CONTROL PROBLEM
No full textThis research paper delves into optimizing the Average Value at Risk (AVaR) using Approximate
Dynamic Programming (ADP) in the context of optimal control problems. The study focuses on
comparing different numerical optimization methods to achieve that. The methods include Bisection,
Gradient Descent, Simulated Annealing, and Conjugate Gradient. The purpose is to assess their
accuracy and computational effectiveness in optimizing AVaR function within discrete time, finite
horizon settings
TRINOMIAL METHOD FOR OPTION PRICING WITH TRANSACTION COSTS FOR SUPPLY CHAIN FINANCING
No full textIn this study, we investigate the trinomial method for option pricing, incorporating trans- action costs into a discrete-time framework. Taking the binomial option pricing model in Cox et al. (1979) as a foundation, we further extend it to construct the trinomial model as referenced in Bjorefeldt et al. (2016). Our research examines the integration of transaction costs into three option pricing models – Black-Scholes, binomial, and trinomial – through the comparative analysis of the results. We also explore the application of the trinomial model as a pricing tool for supply chain financial products, aiming to address financial challenges faced by small and medium-sized businesses. Building upon the case studies outlined by Yun- zhang et al. (2021), we use the framework of American call options. Despite our efforts to integrate our model into the supply chain financing context, we have encountered challenges. Our current model, while proficient in handling fixed parameters, lacks the flexibility required to incorporate the variables needed in supply chain financing scenarios
ADVANCING FORMATIVE USE OF ASSESSMENT IN ADVANCED COMPUTER SCIENCE AT NISS: USING THE RASCH MODEL FOR MOCK EXAMS
No full textAssessment plays a crucial role in shaping education, and this study focuses on enhancing formative assessment practices in Advanced Computer Science at Nazarbayev Intellectual Schools (NIS). The NIS system's incorporation of external assessments for certification purposes has led to modifications in assessment procedures, creating a significant impact on subject structure and content. As students prepare for conclusive external exams, the quality of assessment tools, including mock exams, becomes paramount. However, the absence of statistical testing for the validity and reliability of mock exams in the NIS system raises concerns about assessment accuracy and potential mismatches between teaching and assessment. To address these gaps, this research employs the Rasch model for a secondary analysis of psychometric properties, evaluating reliability, validity, item difficulty levels, and discrimination indices. Utilizing tools like "autopsych: An R Shiny Tool for the Reproducible Rasch Analysis" and "jamovi 2.3.28.0" with the "snowIRT 4.8.8" module, the study aims to enhance the precision of mock exams for a nuanced understanding of student knowledge and skills. Moreover, the research delves into statistical analyses of mock exams in Advanced Computer Science, striving to determine zones of proximal development for individual students. This objective seeks to provide a more accurate measurement of student abilities, thereby facilitating the development of effective approaches for student preparation for summative external assessments.
Overall, 61 students answered CS Paper 1 and CS Paper 2 mock items. Paper 1 had 70 scores from 37 questions, and Paper 2 featured 29 questions with varying difficulties. Employing the Rasch model, the study identified key areas for improvement in the Advanced Computer Science mock exam. Findings highlighted the need for refining question design,
8
scoring systems, and overall structure. Additionally, insights into students' Zone of Proximal Development informed tailored interventions supported by tools like Bloom's Taxonomy and Glaser's Levels of Increasing Competence. Dissemination of these findings equips teachers for effective assessment practices, aligning instructional strategies with students' diverse needs.
The findings not only contribute to the specific context of Advanced Computer Science but also have broader implications for formative assessment practices across subjects and educational contexts. Future studies could explore the longitudinal impact of the Rasch model on mock exams on students' performance in external assessments, contributing to ongoing research in educational assessment
EMPLOYING 2D MATERIAL IN REINFORCING INTERFACE STRENGTH IN FIBER-REINFORCED METAL LAMINATES
No full textThis study was conducted with the aim of improving fiber-metal laminates (FMLs) by
strengthening the interface between layers using 2D materials such as the ɑ-Zirconium
phosphate (ɑ-ZrP). The importance of this topic is emphasized by an analysis of the
literature, which reveals a gap in research in the field of application of ɑ-ZrP
nanomaterials. Even though ɑ-ZrP was discovered and synthesized a long time ago,
there is a lack of studies in using it as a reinforcement for fiber metal laminates. To
achieve this goal a research were conducted to synthesize and characterize ɑ-ZrP
nanoparticles. Then some experiments were performed to find the best composition of
the polymer starting from 0 wt% to 2 wt% of these nanoparticles by using the
commercial polyurethane. Several single lap joint samples using the aluminum sheets
were made and additionally pure polyurethane polymers were examined for the extra
information about the adhesive layer. Mechanical and electrochemical treatments were
implemented for the improvement of the interlocking mechanism between the metal
and adhesive polymer. In addition, the synergetic effect of their combination with
nanoparticles was studied. The behavior of the composite was studied using different
testing types such as the single lap shear strength test and tensile strength test for pure
polyurethane as well as it was analyzed by the Finite Element Analysis (FEA) using the
ABAQUS software. Ultimate tensile strengths were determined for the comparison of
the experimental and numerical parts’ results and the best concentration for the pure
polymer was found to be 1.0 wt% of ɑ-ZrP with 3.5 times enhanced tensile strength,
while for single lap joints it was 0.5 wt% which increased the shear strength by 91.8%.
Mechanical treatment by itself significantly increases the shear strength of lap joints by
35%, although it reduces efficiency in combination with nanoparticles. Conversely,
electrochemical treatment, especially combined with nano reinforcement showed
superior performance in terms of increase of the shear strength by 174% compared to
all other treatments. Although the combination of nanoparticles with both treatments
provided a slight increase in strength by 18.9%, this did not correspond to the
significant increase achieved by electrochemical treatment. The initial hypothesis was
proved and supported with numerical modelin
Electrochemical synthesis of spirocyclic morpholines and tetrahydrofurans via an oxidative dearomatisation strategy
No full textIn search for developing synthetic routes towards spirocyclic morpholines and tetrahydrofurans, we developed a convenient electrochemical protocol for dearomatization of methoxy-benzene derivatives. The electrochemical oxidation of the electron-rich benzene ring followed by intramolecular capture of the cation-radical intermediates accessed desired spirocyclic products. The scope of the reaction demonstrated tolerance of various functional groups such as alkenes, alcohols, bromines, fluorines, trifuoromethyl, cyano, ester and amides with reaction yield varying between 46-91%
IOT TELEGRAM BOT NETWORK DEPLOYMENT AND MEASUREMENTS
No full textThis project was done to resolve connectivity inconsistencies encountered while
transmitting data from devices with temperature and luminosity sensors to a Telegram bot due to instability in WiFi connection. Consequently, Long Range (LoRa) technology was implemented in this project due to its relevance for low-power and wide-area radio communication. The project lifetime lasted from September 2023 to April 2024, in a total of 7 months together with a 1 month break during December and January. Initially esp32 and Pycom LoPy4 were used as devices, however later we switched to both of them to be esp32. Sender device reads data from temperature and luminosity sensors and sends them to the receiver. Aside from that, the receiver device uses OpenWeatherMap API to get the outside temperature in Astana with a given longitude and latitude. Additional functionality has been added to this project like connection to OpenAI API and machine learning implementation. For the OpenAI API part, the receiver device has been connected to the API to access ChatGPT queries and get answers for questions
from it. The ML model implementation uses data gathered from the telegram bot for the past 10 months, which is stored in .csv format. At the end of the day the receiver device connects to OpenWeatherMap API and gets predictions for morning and afternoon temperatures outside of university, using these temperatures and ML model, it posts predicted temperatures for atrium and outside to telegram bot. In the end, multiple tests were conducted with LoRa and other functionality, so that one user uses a sender device and another a receiver device. The connection was tested from different spots, and the main spot for receiver was C4 block in Nazarbayev University, and main spot for sender was green spot in the atrium of university. The connection resulted in a range of -94 to -97 RSSI value which is acceptable for LoRa connection, solving the main problem of the project
USING MICROPHONE AND ML TO DETECT THE PRESENCE OF HUMANS IN SPACE
No full textThis project explores the development and perfor-
mance of a voice recognition system implemented on an ESP32
microcontroller, utilizing a 1D Convolutional Neural Network
(CNN) architecture. The system’s objective is to detect human
presence by recognizing individual vocal characteristics through
real-time audio input. The research extends into quantization
techniques, employing the EON compiler to optimize the CNN
model for efficient execution on the constrained hardware,
reducing memory and flash usage while maintaining accuracy.
The system was evaluated on a dataset split into training and
testing subsets, achieving a remarkable accuracy of 90.72% on
the testing set, surpassing the initial accuracy target of 80% set
during the project’s inception. The integration of the MAX9814
microphone with the ESP32’s Direct Memory Access (DMA)
and built-in I2S protocols enabled high-fidelity audio recording
without delays. This project not only confirms the feasibility of
deploying machine learning models on low-resource microcon-
trollers but also provides a foundation for future enhancements
in biometric-based security and personal identification systems
FACULTY PERCEPTIONS OF ARTIFICIAL INTELLIGENCE (AI) TOOLS IN TEACHING AND LEARNING: A PHENOMENOLOGICAL STUDY
No full textThe use of AI-powered tools, such as ChatGPT, e-learning technologies, and digital platforms with virtual assistants, is becoming increasingly common within educational settings due to globalization and digitalization trends. In an effort to align with the global trends, these advancements are expected to be adopted in Kazakhstan's higher education system as well. Hence, faculty are tasked with implementing these technologies in practice and ensuring student success in using them effectively for their academic needs. The study aims to investigate the faculty perceptions of using AI-driven tools in teaching and learning at higher education institutions in Kazakhstan. To explore the individual perspectives of the faculty towards using AI in education, this study employed a phenomenological design with semi-structured interviews. Seven faculty members from a research-intensive university in North Kazakhstan participated in this research. The findings revealed both positive and negative perceptions of faculty members based on their experience, attitudes, and social influences. AI tools were seen as beneficial for lesson planning, student engagement, feedback, and workload management. The attitudes towards AI within participants’ social circles also influenced the adoption of AI. Nevertheless, concerns were raised about the ethical use, reliability, and potential negative impact on student learning and academic integrity. Additionally, the findings highlighted the importance of clear institutional policies for responsible AI use, as well as the need for training and workshops to enhance faculty and student proficiency in AI within educational settings. Ultimately, the study underscores the importance of collaborative efforts between faculty, institutional administration, and students to facilitate the development of effective policies and training programs that ensure the responsible use of AI technologies in teaching and learning
IMPLEMENTATION OF MACHINE LEARNING METHODS FOR PREDICTING ENERGY DEMAND OF PCM-INTEGRATED BUILDINGS
No full textNumerous machine learning methods have been employed to predict the energy consumption of PCM-integrated buildings. However, the following research gaps have not been addressed yet. 1) Most researchers developed prediction models by only considering building parameters; 2) Only one research team developed prediction models by considering environmental parameters. However, they did not consider some of the important environmental parameters including precipitation and air pressure; 3) No research team have proposed the prediction model by considering the future climate scenario and evaluated the impact of hyperparameters especially for decision tree-based algorithms. This research aims to evaluate the efficacy of different decision tree subcategories (fine, medium, and coarse trees) for predicting energy consumption in PCM-integrated buildings for future climate scenario, considering extensive building and environmental parameters. A database for the energy consumption of PCM-integrated buildings for 11 cities in the hot semi-arid climate zone was created through energy simulations. The results showed that the Fine Decision Tree-based prediction model (FDT3) was the most reliable and accurate prediction model, having R2 values greater than 94% for both training and testing phases. Overall, the developed model will be useful in providing valuable insights into the possibilities for sustainable building design and climate-responsive energy management