Yazar "Yüksek, Ahmet Gürkan" seçeneğine göre listele
Listeleniyor 1 - 11 / 11
Sayfa Başına Sonuç
Sıralama seçenekleri
Öğe An Innovative Approach to Improve Point Location Detection System with ANFIS using RSSI Signals and Fingerprinting Method(Erciyes University, 2024) Yüksek, Emre; Yüksek, Ahmet GürkanLocalization systems have an important place in many areas. GPS (Global Positioning Systems) using data from satellites gives successful results in localization systems. However, localization systems such as GPS, which can be quite successful outdoors, do not achieve the same success indoors because the satellite viewing angle cannot be maintained continuously or due to low reception quality. In this respect, there is a need for localization systems that can provide the most precise localization with the least cost in the interior. This study aims to improve fingerprint-based localization systems, which is a localization method based on Received Signal Strength Indicator (RSSI) data using ANFIS. The proposed system has been shown to give more successful results than the methods frequently used in the literature.Öğe Bilgisayar sistemlerinde veritabanı yönetimi(Cumhuriyet Üniversitesi, 2001) Yüksek, Ahmet Gürkan; Gülcü, AslanÖZET Bu çalışma içerisinde, Veritabanı Yönetim Sistemleri temelleri kısaca incelenmiş ve günümüzde en yaygın olarak kullanılan Internet ortamlarında veritabanı dosyalanna erişmek ve yönetmek (temel işlemler olarak) konusu üzerinde bir uygulama geliştirilmiştir. Geliştirilen uygulamada tanımlanan veritabanı dosyaları üzerinde, belirlenen bir alan yardımıyla kayıt ekleme, listeleme, kayıt silme ve temel arama işlemleri gerçekleştirilmektedir. Burada veritabanı olarak "Access" ve veritabanı yönetim düzeneği olarak da Microsoft Windows İşletim sistemlerinin sağladığı "ODBC" kullanılmıştır. Bu çalışmanın başat amacı bilgisayar sistemlerinin temel kullanım alanlarından olan bilgi ve bilgiyi işleme konulan üzerindeki en gelişmiş düzenek olan veritabanı sistemlerini incelemek ve Internet ortamında bir uygulama geliştirmektir.Öğe Characterization of SiNx grown at different nitrogen flow and prediction of refractive index using artificial neural networks(Elsevier B.V., 2024) Yüksek, Ahmet Gürkan; Horoz, Sabit; Demir, İlkay; Altuntaş, İsmail; Şenadım Tüzemen, EbruSiNx films were grown on silicon substrates by Radio Frequency (RF) magnetron sputtering deposition. The effect of nitrogen flow on the structural and optical properties of the obtained films was investigated using X-ray diffraction, Scanning Electron Microscopy (SEM), UV–Vis–NIR spectrophotometer and spectroscopic ellipsometer, respectively. XRD spectra of the films showed that all films belong to amorphous structure. SEM photographs of SiNx films were analyzed. As a result of the analysis, it was observed that the surfaces of the films had a homogeneous and smooth structure as the nitrogen flow increased. The total and diffuse reflectance spectra of the films were measured and the energy band gaps of the films were determined using the Kubelka-Munk function by using the diffuse reflectance. It was observed that the energy band gap changed as the nitrogen percentage increased. The refractive index of all films was obtained as a function of temperature using a spectroscopic ellipsometer. In the second part of this study, we focused on predicting the temperature dependent refractive indices of the nitrogen flow-dependent films using Artificial Neural Networks (ANN). For the training of the ANN model, wavelength and temperature values from experimental data were used as input and refractive index as output parameters. The simulation and prediction results obtained from this model are compared with the experimental data and interpreted. It is concluded that the ANN approach is suitable for simulating and predicting the temperature dependent refractive index. The models successfully trained with ANN will be especially preferred for predicting the refractive indices of SiNx films, which cannot be measured experimentally, thus providing predictions in non-experimental ranges. In particular, the results obtained by focusing on the ability of the developed artificial neural network (ANN) models to predict the optical properties of SiNx films and their potential to provide information in non-experimental conditions, offer a new approach to quickly and effectively evaluate the optical properties of SiNx films. This approach reveals the importance of artificial intelligence-based methods in materials characterization studies. © 2024Öğe Development of a central controlled automation project on the IoT platform(Peter Lang AG, 2019) Yüksek, Ahmet Gürkan; Arslan, Halil; Çifçi, Gülşah; Elyakan, M. Lemi[No abstract available]Öğe Development of Image Processing Based Line Tracking Systems for Automated Guided Vehicles with ANFIS and Fuzzy Logic(2023) Yüksek, Ahmet Gürkan; Elik, Ahmet UtkuAutomated Guided Vehicles (AGVs) are robotic vehicles with the ability to move using mapping and navigation technologies to perform tasks assigned to them, guided by guides. Using sensor data such as laser scanners, cameras, magnetic stripes or colored stripes, they can sense their environment and move safely according to defined routes. The basic requirement of motion planning is to follow the path and route with minimum error even under different environmental factors. The key factor here is the most successful detection of the guiding structure of a system moving on its route. The proposed system is to equip a mechanical system that can produce very fast outputs and autonomous motion as a result of combining different algorithms with different hardware structures. In the line detection process, the wide perspective image from the camera is designed to be gradually reduced and converted into image information that is more concise but representative of the problem in a narrower perspective. In this way, the desired data can be extracted with faster processing over less information. In this study, the image information is divided into two parts and planned as two different sensors. The fact that the line information was taken from two different regions of the image at a certain distance enabled the detection of not only the presence of the line but also the flow direction. With the fuzzy system, the performance of the system was increased by generating PWM values on two different hardware structures, loading image capture, image processing processes and driving the motors. In order to determine the membership function parameters of the fuzzy system for each input, the ANFIS approach was used on the data set modeling the system. The outputs produced by the ANFIS model were combined into a single fuzzy system with two outputs from the system rules framework and the system was completed. The success of the algorithms was ensured by partitioning the task distribution in the hardware structure. With its structure and success in adapting different technologies together, a system that can be recommended for similar problems has been developed.Öğe Hava kirliliği tahmininde çoklu regresyon analizi ve yapay sinir ağları yönteminin karşılaştırılması(Cumhuriyet Üniversitesi, 2007) Yüksek, Ahmet Gürkan; Karaboğa, DervişSon zamanlarda bölgesel olarak hava kalitesi modellerinin kurulmas ve bu modelleri kullanarak atmosferdeki kirleticilerin younluklarnn ve etkilerinin tahmini ile ilgili birçok bilimsel ara"trmalar yaplmaktadr. Bu tezin amac, hava kirlilii yani havadaki SO2 konsantrasyonu tahmini için Yapay Sinir Alar, Çoklu Dorusal Regresyon Analizi ve Bulank Sinir Alar yakla"mlar ile modellerin kurulmas ve bu modellerden elde edilen sonuçlarn kar"la"trlmasdr. Her yakla"mn kendisine ait avantaj ve dezavantajlar gözlenerek benzer çal"malarda daha uygun olan modelin seçilmesi için bir çal"ma yaplm"tr. Yakla"mlara ait sonuçlar üretebilmek için kullanlan veriler; uygulama alan olarak seçilen Sivas ili "ehir merkezine 1990-2004 yllarna ait hava kirlilii ve meteorolojik verileridir. Bu veriler “Devlet $statistik Enstitüsü”, “Devlet Meteoroloji $"leri Genel Müdürlüü”, “Sivas $l Halk Sal Laboratuarlar” ve daha önce benzer konularda yaplan bilimsel çal"malardan toplanarak hazrlanm" ve özel programlar aracl ile birle"tirilerek bir veri taban olu"turulmu"tur. Veri taban, Sivas "ehri il merkezinde be" dei"ik noktada ölçülen 10 adet meteorolojik ve 2 adet kirlilik parametrelerini içermektedir. Yapay Sinir Alar ile Bulank Sinir Alar için performans en yüksek modellerin belirlenmesi amacyla farkl a yaplar farkl eitim algoritmalar kullanlarak eitilmi" ve çok sayda deneyler gerçekle"tirilmi"tir. Elde edilen sonuçlardan, Yapay Sinir Ann üç model içerisinde en iyi performans sergiledii ve Bulank Sinir Ann Regrasyon Analizinden daha iyi ama Yapay Sinir Alarndan biraz kötü performans gösterdii görülmü"tür. Yapay Sinir Alar sistemlerin geli"tirilmesinde MATLAB program, Klasik Çoklu Dorusal Regresyon Analizi için de STATISTICA program kullanlm"tr. Anahtar kelimeler: Yapay Sinir Alar, Geri Yaylm Alar, Bulank Sinir Alar, Çoklu Dorusal Regresyon Analizi, Hava Kalitesi Modeli.Öğe Modeling of temperature?dependent photoluminescence of GaN epilayer by artificial neural network(Springer, 2023) ŞenadımTüzemen, Ebru; Yüksek, Ahmet Gürkan; Demir İlkay; Horoz, Sabit; Altuntaş İsmailArtificial neural networks (ANNs) are a type of machine learning model that are designed to mimic the structure and function of biological neurons. They are particularly well-suited for tasks such as image and speech recognition, natural language processing, and prediction tasks. The success of an ANN in modeling a particular dataset depends on factors such as the size and quality of the dataset, the complexity of the model, and the choice of training algorithms. High representation rate of a system in the data set can improve the performance of the ANN model. The study we described is focused on using artificial neural networks (ANNs) to model temperature-dependent photoluminescence (PL) characterization of GaN epilayers grown on patterned sapphire substrates (PSS) using the metalorganic chemical vapor deposition (MOCVD) technique. The ANN model is trained using temperature and wavelength as input parameters and intensity as the output parameter, with the goal of accurately predicting the PL intensity of the GaN epilayer as a function of temperature and wavelength. The model is trained using a large set of experimental data and then tested using data that was not presented to the model during training. The results of the study suggest that ANN modeling methodology is an effective and accurate way of modeling temperaturedependent PL of GaN epilayers grown on PSS. The results of the study suggest that ANN modeling methodology can be used to accurately predict the temperature-dependent PL of GaN epilayers grown on PSS. This means that it may be possible to reduce the number of required experimental measurements by using the ANN model to predict PL intensity at different temperatures, based on a smaller set of experimental measurements. This could potentially save time and resources, while still obtaining accurate information about the optical behavior of GaN-based materials at different temperatures.Öğe Predicting optical properties of NiO films fabricated by RF magnetron sputtering: A machine learning approach(Elsevier GmbH, 2025) Yüksek, Ahmet Gürkan; Horoz, Sabit; Altuntaş, İsmail; Demi̇r, İlkay; Tüzemen, Ebru Ş.NiO films with different thicknesses (100, 150, 200, 250, 300 and 400 nm) were grown on glass substrates using the RF Magnetron sputtering method and their optical transmittance properties were analysed with a spectrophotometer. An innovative aspect of this work was the application of machine learning techniques used to derive new insights from experimental data. Four different machine learning algorithms -ANFIS, Artificial Neural Networks (ANN), Support Vector Machines (SVM) and Gaussian Process Regression (GPR)- were tested. While the models were trained using films of different thicknesses, a randomly selected 75 % of the whole dataset was used for model testing and the remaining 25 % of the films were used for testing the models. Among these, ANN and GPR models were found to be the most successful models. Using these models, the energy band gaps were estimated at 1 nm intervals and the values ranged from approximately 3.50 eV to 3.76 eV. © 2024 Elsevier GmbHÖğe Prediction of Wear Properties of Experimental Produced Porcelain Ceramics Using Artificial Neural Networks (ANN)(Sivas Cumhuriyet Üniversitesi, 2023) Yüksek, Ahmet Gürkan; Boyraz, Tahsin; Akkuş, AhmetIn this study, the production and wear properties of porcelain ceramics produced by powder metallurgy method were examined and modelling with artificial neural networks was studied using the experimental data obtained. Porcelain ceramics were prepared by the powder metallurgy route. Mixtures prepared by mechanical alloying method in alumina ball mills were produced by sintering under normal atmospheric conditions after being shaped in a dry press. After drying, the powders were compressed by uniaxial pressing at 200 MPa. The green compacts were sintered at 1100-1200 oC for 1-5 h in air. Then, characterization studies of the sintered samples were carried out and the wear experimental results obtained were converted into data suitable for modelling with artificial neural networks. In the continuation of the study, experimental wear results using artificial neural networks were analysed and modelled. Wear load, wear time, sintering temperature and sintering time were used as artificial neural networks input variables. Wear values were taken as output variables of artificial neural networks. An artificial neural network was established for the prediction of wear properties of porcelain ceramic composites. As a result, the training results and test results were compared with the actual values to control the network performance. A good agreement was observed between the experimental and artificial neural networks model results. After the artificial neural networks estimation, confirmation tests were performed to confirm the experimental results.Öğe Regression Modelling to Study Wear Properties of Experimental Produced Porcelain Ceramics(Sivas Cumhuriyet Üniversitesi, 2023) Yüksek, Ahmet Gürkan; Boyraz, Tahsin; Akkuş, AhmetIn this study, the production and wear properties of porcelain ceramics produced by powder metallurgy method were examined and modelling with regression were studied using the experimental data obtained. Porcelain ceramics were prepared by the powder metallurgy route. Mixtures prepared by mechanical alloying method in alumina ball mills were produced by sintering under normal atmospheric conditions after being shaped in a dry press. After drying, the powders were compressed by uniaxial pressing at 200 MPa. The green compacts were sintered at 1100-1200 oC for 1-5 h in air. Then, characterization studies of the sintered samples were carried out and the wear experimental results obtained were converted into data suitable for modelling with regression. In the continuation of the study, experimental wear results using regression was analysed and modelled. Wear load, wear time, sintering temperature and sintering time were used as regression input variables. Wear values were taken as output variables of regression. An regression was established for the prediction of wear properties of porcelain ceramic composites. As a result, the training results and test results were compared with the actual values to control the network performance. A good agreement was observed between the experimental and regression model results. After the regression estimation, confirmation tests were performed to confirm the experimental results.Öğe Yapay Sinir Ağları Yaklaşımı ile Toprak Kaynaklı Isı Pompasının Performans Analizi(2024) Duman, Netice; Yüksek, Ahmet Gürkan; Caner, Mustafa; Buyruk, ErtanIsı pompaları, binalarda soğutma ve ısıtma için kullanılan konvansiyonel sistemlere verimli ve ulaşılabilir alternatiflerdir. Isı kaynağı olarak toprak ısısını kullanan toprak kaynaklı ısı pompaları (TKIP), ısıtma ve soğutma yüklerini temiz ve sürdürülebilir bir şekilde karşılamak için umut verici teknolojilerdir. TKIP, kurulum ve işletme maliyetleri yüksek olan bir sistemdir. Bu nedenle verimlilik açısından farklı sektörlerde kullanımı uygun olan TKIP sistemini kurmadan performans analizleri yapılabilir olması çok önemlidir. Sistemler kurulmadan önce performans değerlerinin tahmin edilebilecek olduğu modeller ile değerlendirilmesi yaklaşımdan yola çıkılarak, ısı pompası ve sistemin performansı ve yoğuşturucudan atılan ısıyı tahmin etmek için bir yapay sinir ağı (YSA) modeli önerilmektedir. Yapay sinir ağları ile regresyon analizi, girdi ve çıkış verileri arasındaki karmaşık ilişkileri öğrenme yeteneğine sahip bir makine öğrenimi yöntemidir ve problemlerindeki non-lineer ilişkileri etkili bir şekilde modelleyebilir. Sivas ilinde Kurulan deneysel sistem ile ölçülen veriler, YSA'yı eğitmek için eğitim verisi ve test verisi olarak ayrılmıştır ve modelin ilk aşamasında eğitim verisi; ikinci aşamasında ise test verisi kullanılmıştır. Sunulan çalışmada, yatay TKIP’ın performans katsayısını tahmin etmek için çeşitli uygulamalarda kullanılmış ve özellikle sistem modelleme ve sistem tanımlamada yararlı oldukları gösterilmiş yapay sinir ağlarının uygulanabilirliği ortaya konulmuştur. Bu çalışmanın sonucunda, ısı pompası COP R2 değeri 0,9733, TKIP sistemi COP R2 değeri 0,9896 ve yoğuşturucudan atılan ısının YSA modelinin R2 değeri 0,9878 olduğu tespit edilmiştir. Üretilen istatistiksel belirleyiciler üzerinden yola çıkılarak YSA'ların TKIP sisteminde doğru bir yöntem olarak COP tahmini için kullanılabileceği sonucuna varılmıştır.
