Elektrik Elektronik Mühendisliği Bölümü Yayın Koleksiyonu

Bu koleksiyon için kalıcı URI

Güncel Gönderiler

Listeleniyor 1 - 20 / 149
  • Öğe
    Fault ride-through (FRT) capability and current FRT methods in photovoltaic-based distributed generators
    (Elsevier, 2020) Bayrak, Gökay; Ghaderi, Davood; Sanjeevikumar P.
    In this chapter, current methods used for fault ride-through (FRT) capability are examined by researching studies on the capacity of FRT in grid-connected photovoltaic (PV) systems after any failures. These methods have been researched to show the best strategy that can be applied for PV systems by analyzing the FRT methods currently used. FRT methods are classified as external and internal (inverter-resident) methods in the chapter. Energy storage-based methods, flexible alternating current transmission system (FACTS)-based methods, and inverter-resident methods are investigated for PV-based distributed generators. Energy storage-based methods are expensive, and the life cycle of used units is short. Besides, energy storage-based methods are easy to implement. The total cost and complexity of FACTS-based methods were found to be the highest, among others. Regarding grid regulation compatibility, inverter-resident methods are highly effective, and a modified inverter controller method has been found to be the best solution among existing methods.
  • Öğe
    A Demand Side Management Controller Configuration for Interleaved DC-DC Converters Applicable for Renewable Energy Sources
    (Wiley, 2021) Ertekin, Davut; Bayrak, Gökay; Subramaniam, Umashankar
    In Micro-grid applications, accuracy and sensitivity of the Demand Side Management (DSM) process decrease when the load impedance changes. In this study, the impact of the DSM is analyzed and the interleaved structure is presented for DC-DC converter blocks equipped with adaptive PI controllers. This approach reinforces a same voltage source that can be a serial and parallel connection of Photovoltaic (PV) panels in different power rates as the input voltage source to enhance the voltage to the micro-grid DC level and is modeling the power transmission in Renewable Energy Sources (RESs) that they produce limited amounts of power. For the voltage droop problem, the Power-Voltage (P-V) approach is selected. Since the resistive loads are considered in this study, this approach can control DC currents based and depending on the DC voltages in DC micro-grid applications. For per controller block, different values of the gain coefficients are tested and the optimal droop coefficients are presented. All simulations have been done in MATLAB/SIMULINK and a prototype by power around 1kW is tested. The results of the hardware implementation confirm the theoretical and simulation outcomes.
  • Öğe
    A new signal processing-based islanding detection method using pyramidal algorithm with undecimated wavelet transform for distributed generators of hydrogen energy
    (PERGAMON-ELSEVIER SCIENCE LTD, 2022) Yılmaz, Alper; Bayrak, Gökay
    Machine learning-based fault detection methods are frequently combined with wavelet transform (WT) to detect an unintentional islanding condition. In contrast to this condition, these methods have long detection and computation time. Thus, selecting a useful signal processing-based approach is required for reliable islanding detection, especially in real-time applications. This paper presents a new modified signal processing-based islanding detection method (IDM) for real-time applications of hydrogen energy-based distributed generators. In the study, a new IDM using a modified pyramidal algorithm approach with an undecimated wavelet transform (UWT) is presented. The proposed method is performed with different grid conditions with the presence of electric noise in real-time. Experimental results show that oscillations in the acquired signal can be reduced by the UWT, and noise sensitivity is lower than other WT-based methods. The non-detection zone is zero and the maximum detection and computational time is also 75 ms at a close power match.(c) 2022 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.
  • Öğe
    Consideration of graphene material in PCM with aluminum fin structure for improving the battery cooling performance
    (WILEY, 2022) Aslan, Eyyüp; Aydın, Yusuf; Yasa, Yusuf
    Phase change material (PCM) based battery thermal management system (BTMS) provides even heat distribution and lower maximum temperature, but it suffers from low thermal conductivity. In this study, the impact of graphene additive on PCM was analyzed by presenting three experiments with various structures to solve PCM's low thermal conductivity problem. The results demonstrate that there is no positive impact of graphene additive in the first and third structures. The PCM-graphene additive between the second structure's fins significantly improves the battery heat transfer by allowing the battery to cool down 1500 seconds earlier than the graphene-free structure. Moreover, a thermal equivalent circuit model was derived for the second structure because of its enhanced performance. It is shown that the model works accurately and proves its ability to control not only temperature fluctuations but also transient behavior of the battery. This model provides that the battery temperature can be analyzed without experimentation for different charge-discharge scenarios in lithium-ion batteries with a shorter computation time.
  • Öğe
    Cross-entropy method for distribution power systems reconfiguration
    (Wiley, 2019) Sebaa, Karim; Gelen, Ayetül; Nouri, Hassan
    Cross-entropy (CE) is a powerful simulation method for the solution of continuous and combinatory optimization problems. The work presented here utilizes the CE method for the optimal topology of distribution power systems (DPSs). The optimal network switches are determined for the reduction of active power loss. The adapted CE method is tested on three case studies, namely, the 33-node, 83-node, and 880-node DPSs. The results are compared with other reconfiguration algorithms to demonstrate the superiority of the proposed algorithm. The impact of the distributed generation is also investigated. The effective integration of the photovoltaic panels at midday, when their production is highest and meets the peak demand, is showed. Finally, the real-time reconfiguration strategy based on the switching effort reduction is proposed and enhanced via an adequate selection of the initial switch states.
  • Öğe
    Frequency Estimation Methods for Smart Grid Systems
    (Springer, 2018) Mengüç, Engin Cemal; Acır, Nurettin
    Frequency is one of the most significant parameters in the smart grid systems. Thus, accurate frequency estimation becomes an essential task for monitoring, controlling and protecting a real-time smart grid system. In this chapter, we present an overview of the frequency estimation methods in the smart grid system with a focus on real-time adaptive estimation algorithms. Primarily, in Sects. 5.1 and 5.2, the importance of the frequency estimation in the smart grid systems and the challenges encountered in its real-time applications are introduced in detail. In Sect. 5.3, a three-phase power system is then formulated as a two-phase system in the complex domain by using the well-known Clarke’s transformation so as to be able to estimate the frequency of the smart grid system in the real time. For this purpose, the adaptive real-time frequency estimation algorithms are comparatively presented as strictly and widely linear algorithms in Sect. 5.4. The strictly linear algorithms yield optimal solutions only under balanced three-phase systems, whereas the widely linear algorithms give a better solution under both balanced and unbalanced conditions due to the fact that they take into account all statistical information of the system. Considering smart grid applications in real time, the mentioned properties of these algorithms under both balanced and unbalanced conditions are proven in Sect. 5.5.
  • Öğe
    An improved automated PQD classification method for distributed generators with hybrid SVM-based approach using un-decimated wavelet transform
    (Elsevier, 2021) Yılmaz, Alper; Kucuker, Ahmet; Bayrak, Gökay; Ertekin, Davut; Shafie-Khah, Miadreza; Guerrero, Josep M.
    Artificial intelligence (AI) approaches are usually coupled with the wavelet transform (WT) for feature extraction to classify the power quality disturbances (PQDs). Therefore, selecting a useful WT-based signal processing approach is required for a reliable classification, especially in real-time applications. In this study, a new hybrid, un-decimated wavelet-transform (UWT)-based feature extraction method using a support vector machine (SVM) with a "' a trous" algorithm is proposed to classify PQDs in distributed generators (DGs). The proposed method was performed in a real-time application of a DG system to classify PQDs. The derived features were tested on five different machine learning (ML) models by determining the most appropriate classification technique for the proposed UWT-based feature extraction method. An experimental DG system is constituted in the laboratory using a LabVIEW environment, and the proposed method is tested under different grid conditions. Besides, other well-known and studied conventional ML methods were also tested under 25 dB, 30 dB, and 40 dB noise and compared to the developed method. The experimental and simulation results show that the features extracted with the proposed UWT-based method provide much more successful results in classification than the existing wavelet methods in the literature. Furthermore, the proposed method's noise sensitivity performance is much better than other conventional wavelet algorithms, especially in real-time applications.
  • Öğe
    A novel biometric identification system based on fingertip electrocardiogram and speech signals
    (Academic Press, 2021) Guven, Gokhan; Guz, Umit; Gürkan, Hakan
    In this research work, we propose a one dimensional Convolutional Neural Network (CNN) based biometric identification system that combines speech and ECG modalities. The aim is to find an effective identification strategy while enhancing both the confidence and the performance of the system. In our first approach, we have developed a voting-based ECG and speech fusion system to improve the overall performance compared to the conventional methods. In the second approach, we have developed a robust rejection algorithm to prevent unauthorized access to the fusion system. We also presented a newly developed ECG spike and inconsistent beats removal algorithm to detect and eliminate the problems caused by portable fingertip ECG devices and patient movements. Furthermore, we have achieved a system that can work with only one authorized user by adding a Universal Background Model to our algorithm. In the first approach, the proposed fusion system achieved a 100% accuracy rate for 90 people by taking the average of 3-fold cross-validation. In the second approach, by using 90 people as genuine classes and 26 people as imposter classes, the proposed system achieved 92% accuracy in identiying genuine classes and 96% accuracy in rejecting imposter classes.
  • Öğe
    Driver Drowsiness Detection using MobileNets and Long Short-term Memory
    (Institute of Electrical and Electronics Engineers Inc., 2021) Aydemir, Gürkan; Kurnaz, Oguzhan; Bekiryazıcı, Tahir; Avcı, Adem; Kocakulak, Mustafa
    Deep learning has been studied extensively for driver drowsiness detection using video data. However, since the proposed deep learning methods are computationally cumbersome, the commercial driver drowsiness detection methods are still using hand-crafted features such as lane deviation and percentage of eye closure. This study investigates a deep learning model that provides a fair drowsiness detection performance with a lightweight architecture. In the proposed method, Dlib library was used to detect the driver's face in individual frames of video data. The detected faces are fed into a pre-defined convolutional neural network architecture. Then, a long short-term memory network was used to capture the temporal information between the frame sequences to assess the state of drowsiness. The proposed model achieves a detection accuracy of 80% in a popular benchmark dataset. It was also verified that the model could be implemented on a commercial and inexpensive development board with a frame rate of 5 frames per second.
  • Öğe
    Double Compressed AMR Audio Detection Using Spectral Features With Temporal Segmentation
    (Institute of Electrical and Electronics Engineers Inc., 2021) Büker, Aykut; Hanilçi, Cemal
    Double compressed (DC) AMR audio detection is an important audio forensic problem which is used to authenticate the originality of an auido recording. Majority of the existing studies use audio features extracted from the AMR encoder parameters such as linear prediction (LP) coefficients. Recently, we proposed to use the long-term average spectrum (LTAS) features for DC AMR audio detection and promising results were achieved. In this paper, we propose a novel feature extraction techniques which does not require any prior knowledge about the details of the encoding and decoding processes of the AMR codec. The proposed features are extracted from the temporal segmentation of the short-term Fourier transform (STFT) representation of the audio signal. The proposed features are then classified using deep neural network (DNN) classifier. Experimental results conducted on two different databases show that the proposed features considerably outperform the long-term average spectrum (LTAS) features. The average detection rate is improved from 92.44% to 96.48% on MDSVC dataset and from 80.95% to 83.67% on TIMIT database with the proposed features.
  • Öğe
    Design and Validation of an Effective Temperature Compensated-Based FBG Sensor Package for Air Vehicle Applications
    (Institute of Electrical and Electronics Engineers Inc., 2021) Arslan, Mehmet Mücahit; Bayrak, Gökay
    The use of Fiber Bragg Gratings (FBGs) sensors is increasing day by day because of its unique properties such as lightness, immunity to EM and RF signals, minimized harness complexity, and suitability to array manufacturing. Thereby, in the last decades, it has become the first choice in terms of data collection and long-term health monitoring in the aerospace field. On the other hand, like other conventional sensors, Fiber Bragg Grating sensors are also gets affected by the rapid change of environmental conditions such as temperature. In this context, to avoid false alarm situations and compensate for the effect of environmental condition changes, in this study a newly designed temperature compensation metal package structure has been explained and discussed in detail. Obtained results showed that, with the newly designed package structure, the temperature effect had been reduced with success over %95, and the proposed package had been eligible to operate up to 80°C without needing any external reference sensors.
  • Öğe
    An Efficient Full-wave and Switched-capacitor-based AC-DC-DC Converter Configuration; Applicable for High Voltage Gain Industrial Utilizations
    (Institute of Electrical and Electronics Engineers Inc., 2021) Ertekin, Davut; Bilgiç, Mesut Berke; Mutlu, Bülent
    Power electronics circuits are one of the vital and important parts of the industrial applications, where different levels of the DC and AC voltages and currents are necessary to be applied to the different electronic and electrical devices and machines including the test approaches or working under industrial voltages. Meanwhile, many of these devices are designed to work under a certain and fixed industrial level of voltages like 48VDC, 24VDC, 12VDC or 5VDC, but sometimes, another levels of the voltage for a special device can be an issue. Therefore, the converting an AC voltage to a fixed, controllable and robust DC voltage under different level of the load values is necessary. Furthermore, the variation of the input AC voltage also should be considered and the load DC voltages should be hold fixed. This paper uses a Full-wave rectifier circuit to convert the AC voltage to the DC voltage and in the next step this voltage is enhanced and fixed by using a DC-DC step-up converter. Both rectifier and boost converters are high gain since sometimes it is necessary to convert the grid 220VAC to a small DC voltage or enhance a small DC to a higher DC voltage. The proposed converter can be used separately as a high gain AC-DC rectifier or a high gain DC-DC converter, or the combination of the proposed converters can be used to reach a desired DC voltage through an input AC voltage. It can be considered depending on the area where the converter is used. Theoretical analysis is presented and simulation results confirm the correctness of the topology under different working conditions.
  • Öğe
    SA-net: A sequence aware network for the segmentation of the left atrium in cine MRI datasets
    (IEEE Computer Society, 2021) Uslu, Fatmatülzehra; Varela, Marta
    The segmentation of the left atrium (LA) in CINE MRI is a prerequisite for the calculation of LA functional parameters and may be useful when selecting treatments for atrial fibrillation patients. CINE MRI usually captures both the LA and the left ventricle. The similarities between the LA and other cardiac structures complicate the segmentation of the LA and can lead to poor performance of standard 2D segmentation networks. In this study, we present SA-Net, a deep network which implicitly discriminates LA slices from non-LA slices during segmentation, with a sequence modulator using interslice correlations in a global context. Our experiments, conducted on an in-house dataset with 4710-mm thick bSSFP MR image stacks, show that SA-Net leads to good quality segmentation of the LA, with a mean Dice score of 0.89 and a mean Jaccard index of 0.80, outperforming the U-Net.
  • Öğe
    Motor Imagery Signal Classification Using Constant-Q Transform for BCI Applications
    (European Signal Processing Conference, EUSIPCO, 2021) Balim, Mustafa Alper; Hanilçi, Cemal; Acir, Nurettin
    Electroencephalography (EEG) signals have been using for brain-computer interface applications for the last two decades. Motor imagery (MI) signals are one of the EEG signal types formed by imagining a limb's movement. Recently with the help of deep neural networks (DNN) for classifying MI signals using time-frequency (TF) features, considerable performance improvement has been reported. This paper proposes using a well-known TF representation technique called Constant-Q Transform (CQT) for the MI signal classification. Experiments conducted on BCI IV 2b dataset with DNN classifier using CQT spectrogram show that CQT outperforms traditional short-time Fourier transform (STFT) representation.
  • Öğe
    Modified window function for optically thick samples measured by a terahertz time-domain spectroscopic system using a multimode laser diode
    (The Optical Society, 2021) Morikawa, Osamu; Hamada, Dai; Ozturk, Turgut; Yamamoto, Kohji; Kurihara, Kazuyoshi; Kuwashima, Fumiyoshi; Tani, Masahiko
    A low-cost type terahertz time-domain spectroscopic system can be constituted using a multimode laser diode instead of a pulse laser. To suppress noise, a window function is usually used in the Fourier transformation. When this system is used to measure an optically thick sample, the obtained refractive index shows sinusoidal structures. This is caused by ingress of a signal fragment into the measured range and egress of another signal fragment out of the measured range. In addition, a broad positive background component appears in the imaginary part of the refractive index. This is because of the decreased amplitude of the central structure of the signal, which results from the time shift under the window function caused by sample insertion. These false structures can be eliminated by using a modified window function that is shifted with the signal when the sample is inserted.
  • Öğe
    Modelling and Performance Analysis of an Electric Vehicle Powered by a PEM Fuel Cell on New European Driving Cycle (NEDC)
    (Springer Science and Business Media Deutschland GmbH, 2021) Işıklı, Fırat; Sürmen, Ali; Gelen, Ayetül
    Modelling of a complete polymer electrolyte membrane fuel cell (PEMFC) power systems and performance of the models when subjected to common driving cycle are important research issues. In this study a complete PEMFC system, including air and hydrogen supply equipment, fuel cell stack, electrical system and a 75 kW car, is modelled. An efficiency map of a brand new electric motor is directly imported into the model for it. MATLAB & Simulink tools, based on this mathematical model of PEMFC, are used to establish a dynamic model for a vehicle which is electrically supplied by the fuel cell according to cruise characteristics of New European Driving Cycle (NEDC). Model results show significant instabilities during transient operation regarding the late response of the air supply system. Obtained stack characteristics are similar to those obtained in similar studies conducted previously. Performance results of the car based on energy consumption shows perfect agreement with the results of another model developed for an electric vehicle of the same weight and run also on NEDC.
  • Öğe
    LA-Net: A Multi-Task Deep Network for the Segmentation of the Left Atrium
    (Institute of Electrical and Electronics Engineers Inc., 2022) Uslu, Fatmatülzehra; Varela, Marta; Boniface, Georgia; Mahenthran, Thakshayene; Chubb, Henry; Bharath, Anil A.
    Although atrial fibrillation (AF) is the most common sustained atrial arrhythmia, treatment success for this condition remains suboptimal. Information from magnetic resonance imaging (MRI) has the potential to improve treatment efficacy, but there are currently few automatic tools for the segmentation of the atria in MR images. In the study, we propose a LA-Net, a multi-task network optimised to simultaneously generate left atrial segmentation and edge masks from MRI. LA-Net includes cross attention modules (CAMs) and enhanced decoder modules (EDMs) to purposefully select the most meaningful edge information for segmentation and smoothly incorporate it into segmentation masks at multiple-scales. We evaluate the performance of LA-Net on two MR sequences: late gadolinium enhanced (LGE) atrial MRI and atrial short axis balanced steady state free precession (bSSFP) MRI. LA-Net gives Hausdorff distances of 12.43 mm and Dice scores of 0.92 on the LGE (STACOM 2018) dataset and Hausdorff distances of 17.41 mm and Dice scores of 0.90 on the bSSFP (in-house) dataset without any post-processing, surpassing previously proposed segmentation networks, including U-Net and SEGANet. Our method allows automatic extraction of information about the LA from MR images, which can play an important role in the management of AF patients.
  • Öğe
    Deep convolutional neural networks for double compressed AMR audio detection
    (John Wiley and Sons Inc, 2021) Büker, Aykut; Hanilçi, Cemal
    Detection of double compressed (DC) adaptive multi-rate (AMR) audio recordings is a challenging audio forensic problem and has received great attention in recent years. Here, the authors propose to use convolutional neural networks (CNN) for DC AMR audio detection. The CNN is used as (i) an end-to-end DC AMR audio detection system and (ii) a feature extractor. The end-to-end system receives the audio spectrogram as the input and returns the decision whether the input audio is single compressed (SC) or DC. As a feature extractor in turn, it is used to extract discriminative features and then these features are modelled using support vector machines (SVM) classifier. Our extensive analysis conducted on four different datasets shows the success of the proposed system and provides new findings related to the problem. Firstly, double compression has a considerable impact on the high frequency components of the signal. Secondly, the proposed system yields great performance independent of the recording device or environment. Thirdly, when previously altered files are used in the experiments, 97.41% detection rate is obtained with the CNN system. Finally, the cross-dataset evaluation experiments show that the proposed system is very effective in case of a mismatch between training and test datasets.
  • Öğe
    Design of a new UTC-PD to enhance the BW value
    (Institute of Electrical and Electronics Engineers Inc., 2021) Gencal, Huriye; Öztürk, Turgut
    Promising UTC-PDs have been actively used recently in terms of being an alternative to expensive and bulk THz resources used in many areas such as millimeter and Terahertz (THz) signal (wave) production, communication, imaging and spectroscopy. A new product has been developed by modifying the geometric structure of UTC-PD, which has features such as room temperature usage, high operating speed and high operating frequency. The most important feature of the design is that it performs successfully in 0 (zero) Volt, in other words zerobias, applications and has approximately three times more bandwidth than its equivalent.
  • Öğe
    Angular Margin Softmax Loss and Its Variants for Double Compressed AMR Audio Detection
    (Association for Computing Machinery, Inc, 2021) Büker, Aykut; Hanilçi, Cemal
    Double compressed (DC) adaptive multi-rate (AMR) audio detection is an important but challenging audio forensic task which has received great attention over the last decade. Although the majority of the existing studies extract hand-crafted features and classify these features using traditional pattern matching algorithms such as support vector machines (SVM), recently convolutional neural network (CNN) based DC AMR audio detection system was proposed which yields very promising detection performance. Similar to any traditional CNN based classification system, CNN based DC AMR recognition system uses standard softmax loss as the training criterion. In this paper, we propose to use angular margin softmax loss and its variants for DC AMR detection problem. Although using angular margin softmax was originally proposed for face recognition, we adapt it to the CNN based end-to-end DC audio detection system. The angular margin softmax basically introduces a margin between two classes so that the system can learn more discriminative embeddings for the problem. Experimental results show that adding angular margin penalty to the traditional softmax loss increases the average DC AMR audio detection from 95.83% to 100%. It is also found that the angular margin softmax loss functions boost the DC AMR audio detection performance when there is a mismatch between training and test datasets.