Authors:Mustafa ALTIN Abstract: In the present paper, loxodromes, which cut all meridians and parallels of twisted surfaces (that can be considered as a generalization of rotational surfaces) at a constant angle, have been studied in Euclidean 3-space and also some examples have been constructed to visualize and support our theory. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:Muhammed YILDIRIM Abstract: Sleep patterns and sleep continuity have a great impact on people's quality of life. The sound of snoring both reduces the sleep quality of the snorer and disturbs other people in the environment. Interpretation of sleep signals by experts and diagnosis of the disease is a difficult and costly process. Therefore, in the study, an artificial intelligence-based hybrid model was developed for the classification of snoring sounds. In the proposed method, first of all, sound signals were converted into images using the Mel-spectrogram method. The feature maps of the obtained images were obtained using Alexnet and Resnet101 architectures. After combining the feature maps that are different in each architecture, dimension reduction was made using the NCA dimension reduction method. The feature map optimized using the NCA method was classified in the Bilayered Neural Network. In addition, spectrogram images were classified with 8 different CNN models to compare the performance of the proposed model. Later, in order to test the performance of the proposed model, feature maps were obtained using the MFCC method and the obtained feature maps were classified in different classifiers. The accuracy value obtained in the proposed model is 99.5%. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:Üzeyir AYCEL; Yunus SANTUR Abstract: Financial assets considered as time series are chaotic in nature. The main goal of investors is to take a position at the right time and in the right direction by making predictions about the future of this chaotic series. These time series consist of the opening, low, high, and closing prices of a certain period. The approaches used to make predictions about trend direction and strength using moving averages and indicators based on them have noise and lag problems as they are obtained statistically. Candlestick charts, on the other hand, reflect the price-based psychology of bear and bull investors, and facilitate the interpretation of price movements by consolidating the said opening, closing, lowest and highest prices in a single image. It is known that it was applied to Japanese rice markets for the first time in history and there are more than 100 candle patterns. In this study, an extensible architecture software framework using factory patterns and an object-oriented approach is proposed for defining candlestick patterns and developing intelligent learning algorithms based on them. In the studies carried out for financial assets, the profit factor, which shows the portfolio gain of the strategy, is used. It is desirable that this number of wins be greater than 1. When the proposed approach is tested for 5 major financial assets, this value was obtained as greater than 1 for all assets. The proposed software framework can also be used in the development of new robotic approaches in terms of being applicable to all kinds of financial assets in every period. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:Gözde BAYAT; Kazım YILDIZ Abstract: In the last decades, global warming has changed the temperature. It caused an increasing the wildfire in everywhere. Wildfires affect people's social lives, animal lives, and countries' economies. Therefore, new prevention and control mechanisms are required for forest fires. Artificial intelligence and neural networks(NN) have been benefited from in the management of forest fires since the 1990s. Since that time, machine learning (ML) methods have been used in environmental science in various subjects. This study aims to present a performance comparison of ML algorithms applied to predict burned area size. In this paper, different ML algorithms were used to forecast fire size based on various characteristics such as temperature, wind, humidity and precipitation, using records of 512 wildfires that took place in a national park in Northern Portugal. These algorithms are Multilayer perceptron(MLP), Linear regression, Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Decision Tree and Stacking methods. All algorithms have been implemented on the WEKA environment. The results showed that the SVM method has the best predictive ability among all models according to the Mean Absolute Error (MAE) metric. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:Vedat TÜMEN; Kubilay DEMİR Abstract: The use of Unmanned Aerial Vehicles (UAV) in every field is increasing rapidly. In order for UAVs to perform their duties correctly, they must be able to maintain continuous communication with ground stations. The use of WiFi wireless communication protocol has increased due to its high bandwidth. One of the most important threats that can threaten this type of communication is Denial of Service (DoS) attacks. In the event of such an attack, the UAV becomes inaccessible and may crash. Especially when the open port number is known, it becomes much easier to perform attacks that consume the resources of the drone. In this study, a mechanism is proposed to eliminate or at least mitigate the attack risk. This mechanism enables UAVs using wireless communication (WiFi) to communicate using TCP over UDP using middleware. In addition, by periodically changing the UDP open ports with a secret port number sequence known to both parties, it prevents the attacker from using the open port for a long time and renders the attack ineffective. In this study, the effects of the port hopping method on UAVs are evaluated. Test results on real systems shows that the proposed system makes the communication system of UAVs more resistant to DoS attacks by 91.2%. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:Narin ASLAN; Sengul DOGAN, Gonca ÖZMEN KOCA Abstract: Background and Purpose: COVID-19, which started in December 2019, caused significant loss of life and economic losses. Early diagnosis of the COVID-19 is important to reduce the risk of death. Therefore, studies have increased to detect COVID-19 with machine learning methods automatically. Materials and Methods: In this study, the dataset consists of 15153 X-ray images for 4961 patient cases in three classes: Viral Pneumonia, Normal and COVID-19. Firstly, the dataset was preprocessed. And then, the dataset was given to the Cubic Support Vector Machine (Cubic SVM), Linear Discriminant (LD), Quadratic Discriminant (QD), Ensemble, Kernel Naive Bayes (KNB), K-Nearest Neighbor Weighted (KNN Weighted) classification methods as input data. Then, the Local Binary Model (LBP) texture operator was applied for feature extraction. Results: These values were increased from 94.1% (without LBP) to 98.05% using the LBP method. The Cubic SVM method's highest accuracy was observed in these two applications. Conclusions: This study demonstrates that the performance of the presented methods with LBP feature extraction is improved. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:Kübra UYAR; Merve SOLMAZ, Sakir TASDEMIR, Nejat ÜNLÜKAL Abstract: Histology has significant importance in the medical field and healthcare services in terms of microbiological studies. Automatic analysis of tissues and organs based on histological images is an open problem due to the shortcomings of necessary tools. Moreover, the accurate identification and analysis of tissues that is a combination of cells are essential to understanding the mechanisms of diseases and to making a diagnosis. The effective performance of machine learning (ML) and deep learning (DL) methods has provided the solution to several state-of-the-art medical problems. In this study, a novel histological dataset was created using the preparations prepared both for students in laboratory courses and obtained by ourselves in the Department of Histology and Embryology. The created dataset consists of blood, connective, epithelial, muscle, and nervous tissue. Blood, connective, epithelial, muscle, and nervous tissue preparations were obtained from human tissues or tissues from various human-like mammals at different times. Various ML techniques have been tested to provide a comprehensive analysis of performance in classification. In experimental studies, AdaBoost (AB), Artificial Neural Networks (ANN), Decision Tree (DT), Logistic Regression (LR), Naive Bayes (NB), Random Forest (RF), and Support Vector Machines (SVM) have been analyzed. The proposed artificial intelligence (AI) framework is useful as educational material for undergraduate and graduate students in medical faculties and health sciences, especially during pandemic and distance education periods. In addition, it can also be utilized as a computer-aided medical decision support system for medical experts to minimize spent-time and job performance losses. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:Sefa KAZANÇ; Canan AKSU CANBAY Abstract: In this study, the change in the mechanical properties of Niobium (Nb) nanowire with different grain numbers under applied uniaxial tensile deformation was tried to be investigated by Molecular Dynamics (MD) simulation method. The Embedded Atom Method (EAM), which includes many-body interactions, was used to determine the force interactions between atoms. To determine the effect of grain number on the mechanical properties of Nb nanowire, stress-strain curve, young modulus, yield strain and atomic images obtained from the common neighbor analysis method (CNA) were used. It has been determined that necking and breaking of the model nanowire occur at the grain boundaries, however, the number of grains has important effects on the mechanical properties. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:Muhammet Emin ŞAHİN Abstract: The retina layer is the most complex and sensitive part of the eye, and disorders that affect it have a big impact on people's lives. The Optical Coherence Tomography (OCT) imaging technology can be used to diagnose diseases that are caused by pathological alterations in the retina. The importance of early diagnosis in the management of these illnesses cannot be overstated. In this article, an approach based on convolutional neural networks (CNN), a deep learning method, is presented for the detection of retinal disorders from OCT images. A new CNN architecture has been developed for disease diagnosis and classification. The proposed method has been found to have an accuracy rate of 94% in the detection of retinal disorders. The results are obtained by comparing the proposed CNN network model in a deep learning application used in classification with the MobileNet50 network model in the literature. The evaluation parameter values for models trained using the 5-fold cross validation approach for each type of disease in the retinal OCT image dataset are also submitted. The proposed method can clearly be utilized as a decision-making tool to assist clinicians in diagnosing retinal illnesses in a clinical context based on its effectiveness thus far. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:Peter ANTHONY; Betül AY Abstract: With the rising use of facial recognition systems in a range of real-world scenarios and applications, attackers are also increasing their efforts, with a number of spoofing techniques emerging. As a result, developing a reliable spoof detection mechanism is critical. Active-based techniques have been shown to be good at finding spoofs, but they have a number of problems, such as being intrusive, expensive, hard to compute, not being able to be used in many situations, and usually needing extra hardware. This research presented an active-based robust spoof detection technique capable of detecting a wide range of media or 2D attacks while being less intrusive, less expensive, low in complexity, and more generalizable than other active-based techniques. It doesn't require any additional hardware, so it can easily be integrated into current systems. The distortion variations of video frames of the user's face collected at varying distances from the camera are analyzed to detect spoofing. Both the legitimate and spoof attack datasets were created using real-world facial photo and video data. The proposed approach achieved a spoof detection accuracy of 98.18% using both machine learning classifiers and a deep learning model, with an equal error rate and a half total error rate as low as 0.023 and 0.021, respectively. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:Muharrem Tuncay GENÇOĞLU Abstract: Blockchain is one of the most interestingly developing technologies today, with its applications in many fields from smart contracts to cryptocurrencies. In this respect, blockchain is a hot modern topic nowadays. This study presents a mathematical analysis of cryptographic hash functions, which are one of the most important elements for understanding the security foundations of this technology. In this analysis presented; Hash functions, which are one of the building blocks of blockchain technology, used to ensure information integrity, have proven to be resistant to collision resistance, which is very important in data mining. Thus, a theory has been put forward that will contribute to time and energy saving, which is one of the important problems in data mining, in which blockchain technology is used. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:Tülin ÖZTÜRK; Oğuzhan KATAR Abstract: The brain, which consists of nerve cells called neurons, is the center of the nervous system. The rapid and abnormal growth of nerve cells by interacting with each other is called a brain tumor. Undiagnosed or delayed diagnosis of brain tumors lead to death. Although it depends on experience, manually diagnosing and classifying brain tumors is challenging for physicians. Artificial intelligence-based computer systems can help doctors detect brain tumors using the developments in hardware technology and the amount of data increasing daily. This study proposes a deep learning-based system to classify brain MRI images as tumorous or normal using the pre-trained EfficientNet-B0 model. Our radiologist validated a public dataset containing 3000 brain MRI images. The dataset is divided into 70% train, 20% validation, and 10% test. In the test phase after the training, the pre-trained EfficientNet-B0 model achieved high performance with 99.33% accuracy, 99.33% sensitivity, and 99.33% F1 score. In addition, in the evaluation of the test images, the heat maps obtained by the Grad-CAM method were examined by our radiology specialist. The result of evaluations shows that the pre-trained EfficientNet-B0 deep model chooses the right focus areas in its predictions and can be used for clinical tumor detection due to its explainable structure. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:Nursena BAYĞIN; Mehmet KARAKÖSE Abstract: With the developing technology, the production model, which is structured in line with user requests, has become a very popular topic. This production model, which expresses individualization, has become increasingly common. For this reason, it attracts the attention of many researchers and company executives. At this point, studies are concentrated on the concept of mass customization, which expresses personalized production. Considering the related studies, various difficulties are encountered in this production model on issues such as cooperation, trust, and optimization. In this proposed method, a blockchain-based platform is designed to solve the problems of cooperation and trust, one of the most important problems of mass customization. In addition, in this study, the problem of optimization of the production and supply chain process in the manufacturing sector has been examined. This process includes reaching from the producer to the consumer and many parameters. Therefore, the optimization of this process is a very difficult problem. A two-stage system has been proposed to find a solution to this problem. In the first stage, a reliable platform was created by bringing together service providers and buyers in the manufacturing sector with blockchain. In the second stage, the most suitable parties were selected by a genetic algorithm. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:Ecem ÖZEN ÖNER; Muhammed KANCA, Yakup SAY Abstract: In this study, NiMnCoSn alloy was produced in the arc melting furnace and then grounded into small powder particles. After this procedure, particles of alloys were pelletized and heat treatment was applied to pellet alloys for 3 different temperatures (500 oC,700 oC and 900 oC). Differential scanning calorimetry (DSC), X-ray diffraction (XRD) and physical property measuring system (PMMS) were used for determining physical properties of samples. The biggest feature of NiMn-based shape memory alloys is that they are magnetically based. The feature that distinguishes magnetic shape memory alloys from traditional ones is that the shape memory effect is magnetic. For this reason, studies of NiMn-based alloys are becoming very popular. It was observed that, grounding procedure is effected all physical properties of NiMnSnCo shape memory alloys, seriously. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:Fatih BANKUR; Mustafa KAYA Abstract: With its development, artificial intelligence has formed the basis for many studies aimed at facilitating people's lives. More successful results have been tried to be obtained with the increasing data and developing equipment in these studies. It is seen that these developments in artificial intelligence are reflected in the studies related to sign language conversion. In this study, a data set belonging to the letters in the Turkish Sign Language Alphabet was created, and the classification process was carried out with both the deep learning model we created and VGG16, Inceptionv3, Resnet, and Mobilnet models, which are frequently used in image classification. In addition, an open-source data set containing the letters in the American Sign Language Alphabet was organized similar to the data set containing the letters in the Turkish Sign Language Alphabet we created, and Deep Learning models were used to classify the letters in the American Sign Language Alphabet by using this data set. Performance evaluations of the classifications made by Deep Learning Models using both data sets were made. With this study, the results obtained from training Deep Learning methods with different data sets were compared. In addition, it is thought that the study will be useful in determining both the data set and the deep learning method to be used for the studies on the recognition of Sign Language Letters. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:Şule İNCİ; Mehmet AKYÜZ, Sevda KIRBAĞ Abstract: Trametes versicolor (L.) Lloyd known as turkey tail, is a medicinal mushroom belonging to the Polyporaceae. Although the consumption and commercial sale of this mushroom in our country is new, it has been used for centuries as a medicine in some countries, especially in China. In this study, it was aimed to determine the antimicrobial and antioxidant effects of ethanol and methanol extracts of T. versicolor. Its antimicrobial effects were determined by disk diffusion and microdilution method using pathogenic microorganisms such as Escherichia coli, Klebsiella pneumoniae, Pseudomonas aeruginosa, Bacillus megaterium, Staphylococcus aureus, Candida albicans and Trichophyton sp. Total antioxidant level, total oxidant level and DPPH radical scavenging capacity were detected for the antioxidant activity of the mushroom. According to the results obtained, it was seen that these extracts inhibit the growth of microorganisms at different rates (10-21 mm) according to the disk diffusion method. The minimal inhibitory concentrations of T. versicolor against microorganisms used were determined to be between 62.5-250 µg/mL. The TAS and TOS values of the methanol extract were 0.72 mmol Trolox Equiv./L and 18.39, respectively, the TAS and TOS values of the ethanol extract were detected 0.88 mmol Trolox Equiv./L and 16.71 μmol H2O2 Equiv./L, respectively. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:İbrahim Haruna UMAR; Müge Elif ORAKOĞLU FIRAT Abstract: Understanding the physical and mechanical properties of soils subjected to freeze-thaw cycles, including both micro and macrostructures, is critical for achieving the required performance of structures employing it as a structural or support material. An experimental study was carried out on clay soil with varying water content (18%, 21.5%, and 23%) after repeated freeze-thaw cycles (0, 2, 5, 7, 12, and 15). The performance of soil was evaluated using unconfined compressive strength (UCS) and ultrasonic pulse velocity (UPV) tests. The experimental results demonstrated that UCS peak values were observed at the lowest water content before and after the freeze-thaw cycles. The stress-strain curves exhibited strain-softening behavior, and this condition transitioned to strain hardening behavior after freeze-thaw cycles with increment in the water content. Moreover, the highest values of UPV were observed to increase UCS values due to capillary forces at minimum water content. Also, an increase in the number of freeze-thaw cycles resulted in a decrease in the UPV. According to correlations between UPV and UCS values, the highest correlations for water contents were obtained at optimum water content, and a decreasing trend was observed after experiencing a number of freeze-thaw periods. In addition, the Grey Correlation Analysis was performed to show the degree of correlation between the UCS and UPV, water content as well as the freeze-thaw cycles. The results demonstrated that the UPV values have a greater impact on the UCS than other parameters. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:Oktay KARADUMAN; Canan AKSU CANBAY Abstract: Micro/nano scale thin-film shape memory alloys (SMAs) have been used in many different miniaturized systems. Using them as thin-film metal components in fabrication of Schottky photodiodes has started a few years ago. In this work, a new SMA-photodiode device with CuAlNi/n-Si/Al structure was produced by coating nano-thick CuAlNi SMA film onto n-Si wafer substrate via thermal evaporation. The photoelectrical I-V, C-V and I-t photodiode signalization tests were performed under dark and varied artifical light power intensities in room conditions. It was observed that the new device exhibited photoconductive, photovoltaic and capacitive behaviors. By using conventional I-V method, the diode parameters such as electrical ideality factor (n), Schottky barrier height (ϕb) and rectification ratio (RR) of the produced photodevice for the condition of dark environment were computed as 12.5, 0.599 eV and 1266, respectively. As good figure of merits, the photodiode’s performance parameters of responsivity (Rph), photosensivity (%PS) and spesific detectivity (D*) maxima values determined for at -5 V reverse voltage bias and under 100 mW/cm2 of light power intensity condition are as 0.030 A/W (or 30 mA/W), 18693 and 1.33×1010 Jones, respectively. The current conduction mechanism analysis revealed that the space charge limited conduction (SCLC) mechanism is the dominant current conduction mechanism. By the drawn reverse squared C-2-V plots, the values of diffusion potential (Vd), donor concentration (ND), Fermi level (EF) and also barrier height (ϕb) were determined for the SMA-photodiode. The results indicated that the new SMA-photodiode device can be useful in optoelectronic communication systems and photosensing applications. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:Mehmet KUNT; Abdul BAİDAR, Zeynep ŞANLI Abstract: In this work, the concepts of quantum derivative and quantum integral were renamed to be the left quantum derivative and the left definite quantum integral. Symmetrically to the left, a new quantum derivative (the right) and definite quantum integral (the right) were defined. Some properties of these new concepts were investigated and as well as according to do these new concepts some inaccuracies in quantum integral inequalities were corrected. Moreover, some new quantum Hermite-Hadamard type inequalities were established. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:Zeliha TONYALI; Muhammet YURDAKUL, Hasan SESLİ Abstract: Steel braced frame systems (SBFs) having high stiffness and high strength are commonly utilized due to their resistance to lateral seismic forces in regions with high seismicity. In this study, concentrically braced frames (CBFs) having different bracing configurations are used to obtain the significance of the pulse period associated with near-fault (NF) ground motion by time-history dynamic analysis. Besides, far-fault (FF) ground motions are also used to compare with NF ground motion results according to chancing bracing configurations. To achieve dynamic responses of steel frames with different concentric bracings under NF ground motions, which especially have small, medium, and long pulse periods, 3-story and 4-span CBFs having different bracing configurations were selected as an example. 4 FF and 12 NF ground motions having different pulse durations were chosen to evaluate the dynamic response of concentrically braced frames. The results showed that peak ground acceleration (PGA) could be identified as a key parameter that controls the response of braced frames under FF ground motions. In addition, the ratio of the pulse duration to the first mode period is the dominant parameter when this ratio is only greater than 1.0 under the NF ground motions. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:Shahin AHMADOV; Aytuğ BOYACI Abstract: The purpose of this study is to learn how people who speak different languages interpret the same issues, and to compare the results obtained and show the difference between their perspectives. To learn this point of view, we must first turn to open source intelligence. In this execution, a sentiment analysis application was designed using the Python programming language and the Natural Language Processing algorithms in the texts, which were taken as a data set of comments in Azerbaijani, Turkish, Russian and English languages from social media. As the data set, the comments made on 4 subjects: the declaration of Hagia Sophia as a mosque, the objection events that started with the natural gas hike in Kazakhstan, the natural disasters in Turkey, the Ukraine crisis. After loading the texts in four languages from the network environment, after preprocessing, the text was divided into 8 different categories (neutral, fear, joy, anger, sadness, surprise, disgust, shame) by means of the application written in Python programming language based on Data Mining and Machine Learning topics. In the study, precision, sensitivity, accuracy and F1 score were obtained by using Random Decision Forests, K - Near Neighbor Algorithm, Decision Trees, Support Vector Machine, Naive Bayes Algorithm, Logistic Regression, which are machine learning methods. By comparing the results, it was determined that the Logistic Regression method obtained the highest result. A sentiment analysis model was created using the Logistic Regression method, and sentiment analysis was performed for each subject at separation and the results were compared. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:Tayfun GÜNDOĞDU Abstract: In this paper, an interior permanent magnet (IPM) machine having two sets of windings with different number of turns is developed to improve the limited flux-weakening (FW) capability and efficiency, simultaneously. The flux-adjustable range appears to be somewhat limited because of the limited maximum inverter voltage and high magnetic saturation, which degrades the FW capability. To address its restricted FW capability, a unique winding-switching concept is introduced, in which auxiliary coils with lower turns alternately function as the secondary armature winding, resulting in flux-linkage reduction within the same phase. Winding topologies, design considerations, the FW principle, and FW computations have all been addressed. To validate the feasibility of the proposed FW enhancement strategy, a co-simulation procedure based on the 2D finite element method (FEM) and MatLab codes is used to determine the steady-state and FW performance characteristics of IPM machines with various winding topologies. All steady-state and FW performance characteristics of the conventional IPM machine and the proposed IPM machines have been compared quantitatively. Furthermore, to ensure the accuracy of the analytical and numerical calculations provided in this study, the predicted efficiency map of the original Toyota Prius 2010 IPM machine is validated using the efficiency measurements provided. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:Alzubair ALQARAGHULI; Oğuz ATA Abstract: The vehicle detection accuracy and actual in images and videos appear to be very tough and critical duties in a key technology traffic system. Specifically, under convoluted traffic conditions. As a result, the presented study proposes single-stage deep neural networks YOLOv4-3L, YOLOv4-2L, YOLOv4-GB, and YOLOv3-GB. After optimizing the network structure by adding more layers in the right positions with the right amount of filters, the dataset will be repaired and the noise reduced before being sent to the mentoring. This research will be applied to YOLOv3 and YOLOv4. In this study the OA-Dataset is collect and used, the data set is manually labeled with the care of different weathers and scenarios, as well as for end-to-end training of the network. Around the same time, optimized YOLOv4 and YOLOv3 demonstrate a significant degree of accuracy with 99.68 % and precision of 91 %. The speed and detection accuracy of this algorithm are found to be higher than that of previous algorithms. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:Hakan AÇIKGÖZ; Deniz KORKMAZ, Çiğdem DANDIL Abstract: Solar energy systems are increasing their capacity in the energy industry day by day by operating with higher efficiency in parallel with technological developments. The functional operation of photovoltaic (PV) module contributes greatly to the optimal performance of these systems. On the other hand, detection and classification of faults occurring in PV modules are of vital importance in the operation and maintenance of solar energy systems. In this study, the classification of hotspots, which is one of the most common faults in Photovoltaic (PV) modules, is carried out by deep learning methods. First, data augmentation is applied to the images in the training dataset to improve the classification performance. Then, pre-trained deep learning models namely AlexNet, GoogLeNet, ShuffleNet, SqueezeNet, ResNet-50, and MobileNet-v2 are compared on the same test dataset. According to the obtained experimental results, AlexNet has the best performance with an accuracy value of 98.65%, while ResNet-50 provides the worst result with 94.59%. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:Eyüp ERÖZ; Erkan TANYILDIZI Abstract: Multi-objective optimization is a method used to produce suitable solutions for problems with more than one Objective. Various multi-objective optimization algorithms have been developed to apply this method to problems. In multi-objective optimization algorithms, the pareto optimal method is used to find the appropriate solution set over the problems. In the Pareto optimal method, the Pareto optimal set, which consists of the solutions reached by the multi-objective optimization, includes all the best solutions of the problems in certain intervals. For this reason, the Pareto optimal method is a very effective method to find the closest value to the optimum. In this study, the Multi-Objective Golden Sine Algorithm we developed (MOGoldSA), the recently published Multi-Objective Artificial Hummingbird Algorithm (MOAHA), and the Non-Dominant Sequencing Genetic Algorithm II (NSGA-II), which has an important place among the multi-objective optimization algorithms in the literature, are discussed. In order to see the performance of the algorithms on unconstrained comparison functions and engineering problems, performance comparisons were made on performance metrics PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:Haşim PIHTILI Abstract: The most important problem of composite materials, which have been started to be used in every field of technology, is the ability to perform machining operations such as drilling and cutting. In this respect, in this experimental study, its performance in drilling different composite materials with different reinforcements and thicknesses with drills of different types and different diameters was investigated. Three different types of drills were used in drilling operations, namely HSS, TIN and Carbide. Reinforced composites of different types and properties were used as materials. As a result of the drilling processes, the surface roughness of the drilled surface, drilling performance, hole quality and different cutting parameters were examined depending on the rotation and feed rate. The same drilling conditions were applied for each composite material. All result values obtained were transferred to graphs and tables. In addition, photographs of all samples were taken under Scanning electron microscopy (SEM) and the surface roughness of these photographs was examined. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:Aliru Olajide MUSTAPHA; Simeon Gbenga OLADELE, Salihu Folorunsho ADİSA, Yemisi Tokunbo AFOLABİ Abstract: The low-cost feedstocks such as sesame (sesamum indicum) and jatropha (jatropha curcas) seed oils were utilized to optimize the yield of alkyd resins. The experimentally selected input factors ranges in the molar ratios of oil:glycerol (0.3 – 1), phythalic anhydride: glycerol (1 – 3), and catalyst (0.5–1.5 wt. %) for optimization were established using the response surface methodology (RSM) of Box Behken model to improve the alkyd resin yield factors. The optimization solution utilizing CaCO3 catalysts, and a combination of other process factors evaluated, as well as the corresponding desirability functions, was found using analysis of variance (ANOVA) results for refined sesame alkyd resin (RSAR) and refined jatropha alkyd resin (RJAR). The RSAR optimization using a CaCO3 concentration of 1.5 wt. % at a molar ratios of oil:glycerol (1.0:1.0) and phythalic anhydride:glycerol (3.0:1.0), while the RJAR at a similar catalyst concentration of 1.5 wt. %, molar ratio of oil:glycerol (1.0:1.0), and phythalic anhydride:glycrol (2.8:1.0) were observed for the alkyd resin optimization for the two processes. At these reaction conditions, the predicted and experimental biodiesel yield were 48.26 % and 47.29 % for RSAR and 62.07 % and 61.61 % for RJAR, respectively which shows less than 0.5% variations in both cases. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:Emrullah EZBERCİ; Derya AVCI Abstract: The duration of the smart intersection system lights is determined automatically according to the nearest busy. The vehicle at the intersection with the camera is calculated by the image processing process. optimizing the signaling time in traffic signaling. It will be passed to be passed by a system that can be reached later. Also the system can be entered with this remote central management. Manually switch to roads. In this study, it is a smart intersection system used with special permission from Malatya Metropolitan Municipality transportation units. These studies and the benefits they have provided are highlighted. In addition, DARKNET's real-time object detection YOLOV3 deep learning model is used within the scope of in-vehicle real-time traffic system from data images on websites for traffic. The vehicles are placed in the targeted and future-determined database. Positive signaling with information from the designed Process-Based Intersection Management System. Agricultural bounty takes advantage of little stealing gases to be grown to take advantage of time and small items. A clean environment will be created. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Authors:Elif Gözde YILMAZ; Erkut SAYIN, Alper ÖZMEN Abstract: Historical structures, which constitute an important part of our cultural heritage, should be well protected and carried into the future. Masonry arch bridges are significant part of these structures. In this study, the single-span Murat Bey Bridge in the province of Kütahya, built in 1460, was studied as a numerical application. Firstly, three dimensional finite element model of the bridge was constituted with SAP2000 finite element program. Static analysis of the bridge under its own weight was carried out. The modal analysis method was used to obtain the dynamic characteristics of the bridge. Then, time-history analysis method was applied for seismic evaluation of the bridge. For this purpose, the acceleration records of the 1998 Adana, 2003 Bingöl, 2011 Van and 2020 Elazığ earthquakes were taken into consideration. As a result of the dynamic analyses carried out, the displacement and stress graphs occurring on the bridge were examined. The highest displacement and stress values on the historical bridge were obtained from the acceleration records of the 2011 Van earthquake. PubDate: Fri, 30 Sep 2022 00:00:00 +030
Open Access journal
