http://103.99.128.19:8080/xmlui/handle/123456789/27 All collections of E.E.E 2026-09-12T02:48:18Z http://103.99.128.19:8080/xmlui/handle/123456789/586 Title: EEG Signal Classification between Audio and Visual Stimulation and Multi-level Stress Detection Authors: Troyee, Trishita Ghosh; ID:, 21MEE005F Abstract: Stress in the mind and body is a reaction to perceived difficulties or dangers. It can have significant impacts on mental and physical health, affecting cognitive abilities, emotional stability, and physical well-being. To control stress and avoid its long-term impacts, it is essential to comprehend these effects. Mental stress can be detected in many ways and Electroencephalogram (EEG) is one of them. EEG is the graphical representation of Brain's electrical activity. Regular mental stress gives rise to many mental disorders and it may cause various physiological and psychological diseases. As a result, early-stage detection of stress is very important. In this research, brain activity was recorded through EEG headset during inducing different levels of stress from audio-visual stimulus. To better evaluate visual and auditory stress, an automated system is designed to differentiate among various audio and visual evoked potentials. This may further help for designing different assistive devices for the people having visual and hearing disability. Again, many researchers worked on different level of stress detection, but four level stress detection is still unchecked. In this thesis, mental arithmetic tasks were used as both audio and visual stimuli. Here, a framework was proposed to classify different levels of stress in response to audio and visual stimuli and classification was done between these two stimuli by analyzing EEG signals. A proper pre-processing pipeline selection is essential to the effectiveness of EEG-based mental stress detection systems, as noisy EEG data are unlikely to produce improved findings. An ideal EEG pre-processing pipeline was suggested in this work and compared it to the conventional approach now in use. Using the two pre-processing pipelines, raw EEG data was pre-processed which was collected at lab environment. By extracting robust features from the denoised audio and visual data, binary and multi-level stress were classified. Four machine learning (ML) classification models were used for this purpose. On top of that, the audio and visual stimuli was classified with highest 98.37% accuracy using SVM. When the suggested pipeline for pre-processing EEG data was put into practice, a noticeable increase was found in classification accuracy over the traditional approach. With SVM, the best accuracy of 97.14%, 89.01%, 89.59% were achieved for two, three and four level stress detection using visual stimulation. By contrast, the conventional pipeline produced results of 85%, 77.67% and 66.6%. Again, for auditory stimulation, the highest accuracy of 94.51%, 87.7% and 82.63% was found for two, three and four level stress detection using SVM. On the other hand, the traditional pipeline produced results of 90%, 77.46% and 71.69% respectively. Description: A Master of Science (M.Sc) Thesis in Electrical and Electronic Engineering (EEE) Department at Chittagong University of Engineering and Technology (CUET). 2025-05-22T00:00:00Z http://103.99.128.19:8080/xmlui/handle/123456789/585 Title: Techno-Economic Analysis and Design Optimization of Concentrated Solar Power Plants for Sustainable Energy Development in Bangladesh Authors: BHUIYAN, A.B.M. NOUSHAD; ID:, 21MEE025P Abstract: This study assesses the feasibility of Concentrated Solar Power (CSP) in Bangladesh, evaluating four CSP technologies—Solar Tower (ST), Solar Dish (SD), Parabolic Trough Collector (PTC), and Linear Fresnel Reflector (LFR)— across the three highest Direct Normal Irradiance (DNI) regions: Cox’s Bazar, Bandarban, and Chapai Nawabganj. Key performance indicators, including capacity factor, annual energy generation, Levelized Cost of Electricity (LCOE), and Net Present Value (NPV), were analyzed. Among the technologies, PTC showed the best performance, with a 67.5% capacity factor, 889.26 GWh energy generation, and the lowest LCOE (8.98 ¢/kWh) in Chapai Nawabganj. ST had the highest energy yield (472.09 GWh) in Cox’s Bazar with a competitive LCOE (7.49 ¢/kWh). LFR demonstrated strong economic feasibility, achieving an NPV of $382.24 million and a 51.9% capacity factor. SD, while less efficient, was suitable for decentralized applications. A comparative study was conducted against existing CSP projects, including Noor II and Noor III (Morocco) and Dhursar (India), revealing that the designed plants offer higher capacity factors and lower LCOE relative to global benchmarks. The study utilized System Advisor Model (SAM) for performance optimization, considering solar field design and thermal storage integration. Key challenges include high capital costs and land constraints. Recommendations include hybrid CSP-PV systems and optimized heliostat layouts to enhance feasibility. The findings provide valuable insights for policymakers and investors, laying the groundwork for future CSP development in Bangladesh’s transition toward sustainable energy. Description: A Master of Science (M.Sc) Thesis in Electrical and Electronic Engineering (EEE) Department at Chittagong University of Engineering and Technology (CUET). 2025-05-12T00:00:00Z http://103.99.128.19:8080/xmlui/handle/123456789/573 Title: Study of Linear and Non Linear Properties of Photonic Structures via Numerical and Deep Learning Techniques Authors: Rafi, Rakayet; ID, 21MEE008P Abstract: This study focuses on advancing the design, characterization, and optimization of photonic structures by integrating elementary and advanced deep learning tech niques. The primary aim is to develop predictive models that enhance both the accuracy and efficiency of photonic structure simulations, enabling faster design it erations and real-time optimization.This is aimed to substitute the conventionally used simulation softawares for similar purpose.The first part involves utilizing Ar tificial Neural Networks (ANNs) to predict key linear optical parameters of PCFs, including effective refractive index, dispersion, mode purity, and effective area. The ANN model achieved a high prediction accuracy, evidenced by a Mean Squared Error (MSE) of 0.0218, with a training time of approximately 31 seconds. This rep resents a substantial reduction in computational time compared to traditional nu merical methods while maintaining precise predictions, demonstrating the model’s practical applicability.The second objective focuses on modeling the nonlinear dy namics of supercontinuum generation in optical fibers, with particular attention to chalcogenide planar waveguides. An ANN-based deep learning model was em ployed to accurately capture the complex pulse propagation dynamics, achieving an MSE of 0.0032 with a computational training time of 472 seconds. The model’s abil ity to predict spectral evolution across various pump powers accurately provides a significant potential for real-time optimization of experimental setups.The third objective introduces a novel approach using Closed-Form Continuous-Time Neural Networks (CfC) for predicting supercontinuum spectra in waveguides with diverse material compositions. The CfC model outperformed conventional methods such as Recurrent Neural Networks (RNN) and Long Short-Term Memory (LSTM) net works, achieving high accuracy with MSE values as low as 7.4199 × 10−4 for a three material dataset and 6.9408 × 10−4 for a four-material dataset. The model demon strated low computation time, confirming its robustness, scalability, and adaptabil ity across different materials.Overall, this research demonstrates that integrating machine learning techniques with photonics significantly enhances the accuracy of predictive models while reducing computational time. These advancements have practical implications for the design, optimization, and real-time control of photonic devices, paving the way for future innovations in the field. Description: A master of engineering thesis in EEE department at CUET 2024-10-29T00:00:00Z http://103.99.128.19:8080/xmlui/handle/123456789/544 Title: A MULTI BAND POWER OSCILLATION DAMPING CONTROLLER for INTERCONNECTED HYBRID AC MICRO-GRIDS SYSTEM Authors: RUMKY, TAJRIN JAHAN Abstract: The increasing integration of renewable energy sources (RESs) with conventional energy sources units in a hybrid AC microgrid system has led to the emergence of power oscillation damping (POD) controller for damping low frequency oscillations (LFOs). A local microgrid has limited energy generation capacity and may need to resort to load shedding during a power shortage. Interconnected hybrid AC microgrid systems are essential for enhancing the reliability and stability of power supply by allowing mutual support during outages or power deficiencies. However, the integration of these subsystems often results in oscillatory modes that could compromise stability and operational efficacy. This thesis presents the design and implementation of a multi-band power oscillation damping (MB-POD) controller implemented for an interconnected hybrid AC microgrid system. The proposed MB-POD controller aims to mitigate LFOs across various frequency bands, enhancing the overall stability and reliability of the microgrid. LFOs are caused by integration of generation sources, variety of dynamic load scenarios, and load disturbances in any of the systems. These oscillations could take place locally or between microgrids. This study introduces an energy storage system (ESS)-based POD designed to mitigate such instabilities in an interconnected AC microgrid system. This study also elucidates the complex dynamics of power oscillation in an interconnected AC hybrid microgrid system, highlighting the deficiencies in conventional power oscillation damping methods. Utilizing advanced mathematical models and simulation techniques, we propose a novel approach to dampen the oscillatory modes effectively. Using the proposed multi-band damping controller as well as the extant single-band power oscillation damper (SB-POD), interconnected hybrid AC microgrid system is intended to operate in a MATLAB/Simulink environment. Using time-domain simulations the v proposed controller's performance is evaluated. Simulations and case studies elucidate the damper's efficacy in enhancing system stability while optimizing power flow and reducing the transient response time. The findings indicate substantial improvements in damping multiple oscillatory modes by improving the damping ratio from 9.1% and reducing oscillations by approximately 4-6%, making it a promising solution for modern power systems. MB-POD than SB POD across various microgrids, thus paving the way for more resilient and adaptive interconnected hybrid AC microgrid system. Description: A Master of Science (M.Sc) Thesis in Electrical and Electronic Engineering (EEE) Department at Chittagong University of Engineering and Technology (CUET). 2024-10-01T00:00:00Z