Yazar "Durgut, Rafet" seçeneğine göre listele
Listeleniyor 1 - 20 / 22
Sayfa Başına Sonuç
Sıralama seçenekleri
Öğe Adaptive binary artificial bee colony algorithm(Elsevier, 2021) Durgut, Rafet; Aydin, Mehmet EminMetaheuristics and swarm intelligence algorithms are bio-inspired algorithms, which have long standing track record of success in problem solving. Due to the nature and the complexity of the problems, problem solving approaches may not achieve the same success level in every type of problems. Artificial bee colony (ABC) algorithm is a swarm intelligence algorithm and has originally been developed to solve numerical optimisation problems. It has a sound track record in numerical problems, but has not yet been tested sufficiently for combinatorial and binary problems. This paper proposes an adaptive hybrid approach to devise ABC algorithms with multiple and complementary binary operators for higher efficiency in solving binary problems. Three prominent operator selection schemes have been comparatively investigated for the best configuration in this regard. The proposed approach has been applied to uncapacitated facility location problems, a renown NP-Hard combinatorial problem type modelled with 0-1 programming, and successfully solved the well-known benchmarks outperforming state-of-art algorithms. (C) 2020 Elsevier B.V. All rights reserved.Öğe Adaptive binary artificial bee colony for multi-dimensional knapsack problem(Gazi Univ, Fac Engineering Architecture, 2021) Durgut, Rafet; Aydin, MehmetThe efficiency and effectiveness of metaheuristic optimization algorithms is managed with diverse search and fast approximation in the solution space. A balanced exploration and exploitation capability is required to achieve by the neighborhood operators towards the aimed efficiency. The majority of metaheuristic algorithms use either single operator or limited to genetic operators, which impose serious boundaries upon performance. In order to avoid this limitation, multiple neighborhood operators can be used within the search process orchestrated by a selection scheme. In this study, an adaptive operator selection scheme is studied with multiple binary operators embedded within artificial bee colony algorithm to solve the multidimensional knapsack problem (MKP) as a renown NP-Hard combinatorial problem. It is implemented for modelling and solving many real-world problems, while it is not trivial to offer a good solution within a reasonable timeframe. A parametric study has been conducted for the approach proposed in this study. The success of the proposed approach has been demonstrated and discussed with comparative analysis using three different classes of benchmark problem sets.Öğe Analyzing the performances of evolutionary multi-objective optimizers on design optimization of robot gripper configurations(Tubitak Scientific & Technological Research Council Turkey, 2021) Dorterler, Murat; Atila, Umit; Durgut, Rafet; Sahin, IsmailRobot grippers are widely used in a variety of areas requiring automation, precision, and safety. The performance of the grippers is directly associated with their design. In this study, four different multiobjective metaheuristic algorithms including particle swarm optimization (MOPSO), artificial algae algorithm (MOAAA), grey wolf optimizer (MOGWO) and nondominated sorting genetic algorithm (NSGA-II) were applied to two different configurations of highly nonlinear and multimodal robot gripper design problem including two objective functions and a certain number of constraints. The first objective is to minimize the difference between minimum and maximum forces for the assumed range in which the gripper ends are displaced. The second objective is force transmission rate that is the ratio of the actuator force to the minimum holding force obtained at the gripper ends. The performance of the optimizers was examined separately for each configuration by using pareto-front curves and hyper-volume (HV) metric. Performances of the optimizers on the specific problem were compared with results of previously proposed algorithms under equal conditions. With respect to these comparisons, the best-known results of the configurations were obtained. Furthermore, the pareto optimal solutions are thoroughly examined to present the relationship between design variables and objective functions.Öğe Channel selection and feature extraction on deep EEG classification using metaheuristic and Welch PSD(Springer, 2022) Cizmeci, Huseyin; Ozcan, Caner; Durgut, RafetBrain computer interfaces are important for different application domain such as medical, natural interfaces and entertainment. Besides the difficulty of gathering data from the human brain via different channel probs, preprocessing of data is another different and important task that must be solved in order to get better achievement. Selection of the most active channels is an important problem to achieve high classification accuracy. Metaheuristics are good solutions for selecting the optimal subset from the original set, as they have the ability to obtain an acceptable solution in a reasonable time. At the same time, it is necessary to use the correct feature extraction method so that the data can be properly represented. In addition, traditional deep learning methods used for emotion recognition ignore the spatial properties of EEG signals. This reduces the classification accuracy. In this study, we used artificial bee colony optimization algorithm on the seed dataset to increase the classification accuracy. We implemented and tested four different variations of this algorithm. Then, we extracted the features of the obtained channels with the Welch PSD method. We used enhanced capsule network as a machine learning algorithm and showed the best configuration to solve the problem. At the end of the process, 99.98% training and 99.83% test accuracy rates were obtained.Öğe A Comparative Analysis for Binary Search Operators used in Artificial Bee Colony(Ieee, 2021) Atli, Ibrahim; Durgut, Rafet; Aydin, Mehmet EminMetaheuristic optimization algorithms are developed to find the best or near-best solutions within a reasonable time frame utilizing various neighbourhood functions (i.e., operators). Variety of studies have been proposed for structural modifications on metaheuristic approaches or utilization of various operators. Some of these operators help fast convergence at the beginning but lose efficiency relatively or completely towards the end or vice versa. The individual and collective behaviors of operators in the search space plays crucial role in producing fruitful solutions to approximate the optimum and to devise useful adaptive selection schemes in the cases of using multiple operators. To the best of our knowledge, collective behaviour of binary operators has not been analysed comprehensively. In this study, the characteristics and collective behaviour of operators that can work on discrete decision variables within an artificial bee colony are investigated over solving OneMax and SUKP problems utilizing 9 different operators. The results conclude that disABC, GBABC and twoOptABC operators are more effective in solving OneMax problems, while GBABC and twoOptABC are more effective (especially towards end) in the SUKP problems.Öğe A comprehensive investigation into the performance of optimization methods in spur gear design(Taylor & Francis Ltd, 2020) Atila, Umit; Dorteder, Murat; Durgut, Rafet; Sahin, IsmailThe design of gears with a minimum weight is an optimization problem that has been widely discussed in the literature. Various recent metaheuristic optimization methods, along with the conventional methods, have produced successful results for design optimization problems. In this study, a comprehensive investigation was conducted into the solution of the spur gear design problem in metaheuristic optimization methods. The artificial algae algorithm, artificial bee colony and whale optimization algorithm were applied to the problem for the first time. The grey wolf optimizer and particle swarm optimization were also applied. The results were compared with the performance of the genetic algorithm, simulating annealing and particle swarm optimization, applied in previous studies. A statistical evaluation of these methods applied under the same conditions was carried out in terms of stability. It was shown that the new methods demonstrated significantly improved performance in solving the gear design problem compared to existing methods.Öğe Estimation of wear performance of AZ91 alloy under dry sliding conditions using machine learning methods(Elsevier, 2021) Aydin, Fatih; Durgut, RafetThe wear behavior of AZ91 alloy was investigated by considering different parameters, such as load (10-50 N), sliding speed (160-220 mm/s) and sliding distance (250-1000 m). It was found that wear volume loss increased as load increased for all sliding distances and some sliding speeds. For sliding speed of 220 mm/s and sliding distance of 1000 m, the wear volume losses under loads of 10, 20, 30, 40 and 50 N were calculated to be 15.0, 19.0, 24.3, 33.9 and 37.4 mm(3), respectively. Worn surfaces show that abrasion and oxidation were present at a load of 10 N, which changes into delamination at a load of 50 N. ANOVA results show that the contributions of load, sliding distance and sliding speed were 12.99%, 83.04% and 3.97%, respectively. The artificial neural networks (ANN), support vector regressor (SVR) and random forest (RF) methods were applied for the prediction of wear volume loss of AZ91 alloy. The correlation coefficient (R-2) values of SVR, RF and ANN for the test were 0.9245, 0.9800 and 0.9845, respectively. Thus, the ANN model has promising results for the prediction of wear performance of AZ91 alloy.Öğe Gesture Recognition using SAX Method(Ieee, 2016) Kurnaz, Ismail; Durgut, RafetIn this study, an application is developed to recognize human gestures using data which was recorded by using Microsoft Kinect. The data set used in the study is MSRC-12, and it is created by Microsoft. It has several daily human gestures which were recorded from different users. Before gesture recognition process, recorded data was reduced by PAA method and then it was classified by SAX method. Symbols (which are generated by SAX) of percentage similarity is calculated by developed algorithm. The application can recognize all human gestures in dataset correctly.Öğe Global distribution center number of some graphs and an algorithm(Edp Sciences S A, 2019) Durgut, Rafet; Kutucu, Hakan; Turaci, TufanThe global center is a newly proposed graph concept. For a graph G = (V(G), E(G)), a set S subset of V(G) is a global distribution center if every vertex v is an element of V(G)\S is adjacent to a vertex u is an element of S with |N[u] boolean AND S| >= |N[v] boolean AND (V(G)\S)|, where N(v) = {u is an element of V(G)|uv is an element of E(G)} and N[v] = N(v) ? {v}. The global distribution center number of a graph G is the minimum cardinality of a global distribution center of G. In this paper, we investigate the global distribution center number for special families of graphs. Furthermore, we develop a polynomial time heuristic algorithm to find the set of the global distribution center for general graphs.Öğe A heuristic algorithm to find rupture degree in graphs(Tubitak Scientific & Technological Research Council Turkey, 2019) Durgut, Rafet; Turaci, Tufan; Kutucu, HakanSince the problem of Konigsberg bridge was released in 1735, there have been many applications of graph theory in mathematics, physics, biology, computer science, and several fields of engineering. In particular, all communication networks can be modeled by graphs. The vulnerability is a concept that represents the reluctance of a network to disruptions in communication after a deterioration of some processors or communication links. Furthermore, the vulnerability values can be computed with many graph theoretical parameters. The rupture degree r(G) of a graph G = (V, E) is an important graph vulnerability parameter and defined as r(G) = max{omega(G - S) - vertical bar S vertical bar - m(G - S) : omega(G - S) >= 2, S subset of V}, where omega(G - S) and m(G - S) denote the number of connected components and the size of the largest connected component in the graph G - S, respectively. Recently, it has been proved that finding the rupture degree problem is NP- complete. In this paper, a heuristic algorithm to determine the rupture degree of a graph has been developed. Extensive computational experience on 88 randomly generated graphs ranging from 20% to 90% densities and from 100 to 200 vertices shows that the proposed algorithm is very effective.Öğe Human Gesture Recognition using Keyframes on Local Joint Motion Trajectories(Science & Information Sai Organization Ltd, 2017) Durgut, Rafet; Findik, OguzHuman Action Recognition (HAR) systems are systems that recognize and classify the actions that users perform against the sensor or camera. In most HAR systems, an input test data is compared with the reference data in the database using various methods. Classification process is performed according to the result obtained. The size of the test or reference data directly affects the operation speed of the system. Reduced data size allows a significant performance increase in system operation speed. In this study, action recognition method is proposed by using skeletal joint information obtained by Microsoft Kinect sensor. Splitting keyframes are obtained from the skeletal joint information. The keyframes are observed as a distinguishing feature. Therefore, these keyframes are used for the classification process. Keeping the keyframes instead of keeping the position or angle information of action in the reference database can benefit from memory and working time. The weight value of each keyframes is calculated in the method. The problem of temporal differences that occur when comparing test and reference action is solved by Dynamic Time Warping (DTW). The k-nearest neighbor's algorithm is used for classification according to the obtained results from DTW. The sample has been tested in a data set so that the success of the method can be tested. As a result, 100% correct classification was achieved. It is also suitable for working at real time systems. Breakpoints can also be used to provide feedback to the user as a result of the classification process. The magnitude and direction of the keyframes, the change in the trajectory of joint, the position and the time of its existence also give information about the time errors.Öğe Improved binary artificial bee colony algorithm(Zhejiang Univ, 2021) Durgut, RafetThe artificial bee colony (ABC) algorithm is an evolutionary optimization algorithm based on swarm intelligence and inspired by the honey bees' food search behavior. Since the ABC algorithm has been developed to achieve optimal solutions by searching in the continuous search space, modification is required to apply it to binary optimization problems. In this study, we modify the ABC algorithm to solve binary optimization problems and name it the improved binary ABC (IbinABC). The proposed method consists of an update mechanism based on fitness values and the selection of different decision variables. Therefore, we aim to prevent the ABC algorithm from getting stuck in a local minimum by increasing its exploration ability. We compare the IbinABC algorithm with three variants of the ABC and other meta-heuristic algorithms in the literature. For comparison, we use the well-known OR-Library dataset containing 15 problem instances prepared for the uncapacitated facility location problem. Computational results show that the proposed algorithm is superior to the others in terms of convergence speed and robustness. The source code of the algorithm is available at .Öğe Iskelet bilgisi üzerinde ağırlıklı dinamik zaman bükmesi ve sembolik birleştirme yaklaşımı metotları kullanarak yeni bir hareket tanıma sistemi(2017) Durgut, Rafet; Kurnaz, İsmailSensörler ile donatılmış derinlik kamera cihazlarının maliyetlerinin ekonomik olması nedeniyle, günümüzde kullanım alanları artmakta ve yaygınlaşmaktadır. Bu çalışmada bu tür cihazların en çok kullanılanlarından biri olan Kinect cihazından elde edilen veriler üzerinde, Ağırlıklı Dinamik Zaman Bükmesi ve Sembolik Birleştirme Yaklaşımı yöntemleri birlikte kullanılarak yeni bir hareket tanıma yöntemi geliştirilmiştir. Geliştirilen yöntem günlük hareketlerin yer aldığı veri setinde test edilmiş ve %98.15 oranında bir başarı ile günlük hareketler tanınabilmiştirÖğe Küme birleşimli sırt çantası probleminin adaptif yapay arı kolonisi algoritması ile çözümü(2021) Durgut, Rafet; Yavuz, Ilim Betül; Aydın, MehmetMeta-sezgisel ve sürü zekâsı algoritmaları, NP-Zor optimizasyon problemlerine yaklaşık çözümler sunmak için uzun süredir kullanılmaktadır. Özellikle kombinatoryal ve ikili problemler söz konusu olduğunda, algoritmalar içerisine gömülü komşu çözüm üretmek için kullanılan operatör fonksiyonları, aramanın çeşitliliğine sınırlamalar getirirken algoritmaların başarısında önemli bir rol oynar. Bu tür sınırlamalardan kaçmak ve çeşitliliği iyileştirmek için, birden fazla operatörün tek bir operatör yerine bir seçim şeması yoluyla kullanılması tercih edilir. Daha önce farklı sürü zekâsı ve meta-sezgisel algoritmalarla çeşitli kombinatoryal problemleri çözmek için bir dizi operatör seçim şeması kullanılması daha yüksek etkinlik elde etmek için kullanılmıştır. Bu makalede, küme birleşimli sırt çantası problemleri, ilk kez, alternatif operatör seçim şemaları aracılığıyla seçilen birden fazla operatör içeren ikili bir yapay arı kolonisi algoritması ile çözülmüştür. Önerilen yöntem için farklı kredi atama yaklaşımları, farklı kayan pencere boyutları ve parametre konfigürasyonları test edilmiştir. Seçim şemalarının özellikleri kapsamlı olarak 30 kıyaslama problemi üzerinde incelenmiştir. Bu problem kümeleri için en iyi performans gösteren algoritma konfigürasyonu önerilmiştir. Çalışma, başarılı bir seçim şemasına sahip adaptif ikili yapay arı kolonisi algoritmasını sunmaktadırÖğe On topological properties of some molecular cactus chain networks and their subdivisions by using line operator(Taylor & Francis Ltd, 2022) Turaci, Tufan; Durgut, RafetThe mathematical chemistry is the part of theoretical chemistry which is concerned with applications of mathematical applications and methods to chemical problems. Graph theory is the most important part of mathematical chemistry. It studies of descriptors in quantitative structure property relationship (QSPR) and quantitative structure activity relationship (QSAR) studies in the chemistry science. Let G = (V(G), E(G)) be a chemical graph without directed and multiple edges and without loops. There are a lot of topological indices in QSPR/QSAR studies. In this paper, some degree-based topological indices namely first general Zagreb index, general Randic connectivity index, general sum-connectivity index, atom-bond connectivity index, geometric-arithmetic index, ABC (4)(G) index and GA (5)(G) index are computed for the line graphs and para-chain graphs of meta-chain M-n , para-chain L-n and ortho-chain O-n .Öğe Parametrik değerleri belirlenen merminin deforme olabilen yüzeylere verdiği zararın modellenmesi ve bilgisayar grafikleri ile simulasyonu(Karabük Üniversitesi, 2013) Durgut, Rafet; Kurnaz, İsmailBu çalışmada; bilgisayar grafikleri kullanılarak, parametrik değerleri bilinen mermilerin deforme olabilen yüzeylere verdiği zararın bilgisayar grafikleri ile modellenmesi ve benzetimi gerçekleştirilmiştir. Hafif silah mühimmatı türündeki mermi için kullanılan parametrik değerler; ilk çıkış hızı, kütlesi, sürtünme katsayısı ve boyutudur. Bu çalışmada, parametrik değerleri birbirlerinden farklı olan 5 mermi çeşidinin atış sonrası parametrik özellikleri bilinen 5 farklı malzeme yüzeyi üzerinde oluşturdukları hasarlar gösterilmiştir. Mermi türü ve yüzey çeşidi sayısı parametrik değerleri girilmesi koşulu ile arttırılabilmektedir. Uygulamada merminin, yüzeye vereceği hasar, hem sayısal hem de görsel olarak gösterilmektedir. Uygulamada yer alan 3B öğeler Blender ile modellenip, anime edilmiştir. Malzemelerin parametrik değerleri ANSYS yazılımından alınmıştır. Benzetim işleminin gösterildiği uygulama OGRE3D kütüphanesi kullanılarak C++ programlama dili ile hazırlanmıştır.Öğe Performance enhancement of automatic voltage regulator by modified cost function and symbiotic organisms search algorithm(Elsevier - Division Reed Elsevier India Pvt Ltd, 2018) Celik, Emre; Durgut, RafetThis article attempts to solve the problem of efficient design of proportional + integral + derivative (PID) controller applied to popular automatic voltage regulator (AVR) system by employing recently introduced symbiotic organisms search (SOS) algorithm, for the first time. PID controller design needs proper determination of three control parameters. Such a design problem can be taken as an optimization task and SOS is invoked to find out better controller parameters through a new cost function defined in the paper, which allows to evaluate the control behavior in both time-domain and frequency-domain. For the performance analysis, distinct analysis techniques are deployed such as transient response analysis, root locus analysis and bode analysis. Besides, robustness analysis of the closed-loop control system tuned by SOS is performed with regard to parameter uncertainties and external disturbance. The efficacy of the presented technique is widely illustrated by comparing the obtained results with those reported in some prestigious journals and it is shown that our proposal leads to a more satisfactory control performance from the perspective of both time-domain and frequency-domain specifications while with a good robustness to parameter uncertainties and unknown changes in the system output. (C) 2018 Karabuk University. Publishing services by Elsevier B.V.Öğe Prediction of wear performance of ZK60/CeO2 composites using machine learning models(Elsevier Sci Ltd, 2023) Aydin, Fatih; Durgut, Rafet; Mustu, Mustafa; Demir, BilgeIn this study, ZK60 magnesium matrix composites were produced with different content of CeO2 (0.25, 0.5 and 1 wt%) by hot pressing. The wear behaviour of the samples was investigated under loads of 5 N, 10 N, 20 N and 30 N, at sliding speeds of 75 mm/s, 110 mm/s and 145 mm/s. The worn surfaces, wear debris, and counterface material was analysed to reveal the wear mechanisms. Five machine learning algorithms were established to compare their prediction abilities of wear behaviour on a limited dataset measured under different test operations. The hyperparameter tuning phase of each model was conducted to provide a fair comparison. The prediction results were examined under various statistical measures. In the light of prediction results, the superior model was determined.Öğe Reinforcement Learning-Based Adaptive Operator Selection(Springer International Publishing Ag, 2021) Durgut, Rafet; Aydin, Mehmet EminMetaheuristic and swarm intelligence approaches require devising optimisation algorithms with operators to let produce neighbouring solutions to conduct a move. The efficiency of algorithms using single operator remains recessive in comparison with those with multiple operators. However, use of multiple operators require a selection mechanism, which may not be always as productive as expected; therefore an adaptive selection scheme is always needed. In this study, an experience-based, reinforcement learning algorithm has been used to build an adaptive selection scheme implemented to work with a binary artificial bee colony algorithm in which the selection mechanism learns when and subject to which circumstances an operator can help produce better and worse neighbours. The implementations have been tested with commonly used benchmarks of uncapacitated facility location problem. The results demonstrates that the selection scheme developed based on reinforcement learning, which can also be named as smart selection scheme, performs much better that state-of-art adaptive selection schemes.Öğe Sanal gerçeklik kullanarak hareket tanıma temelli fizik tedavi ve rehabilitasyon uygulamasının geliştirilmesi(Karabük Üniversitesi, 2018) Durgut, Rafet; Fındık, Oğuzİnsan makine etkileşimi, insan hayatını kolaylaştırmayı amaçlayan yöntemlerin geliştirildiği, disiplinlerarası bir çalışma alanıdır. Bu alanda yapılan araştırmalar ile çeşitli disiplinlerde karşılaşılabilecek problemlere yaklaşımlar getirilerek, insan yaşamına olumlu yönde katkı sunmak amaçlanmaktadır. Özellikle, Fizik Tedavi ve Rehabilitasyon alanında insan makine etkileşiminin uygulanabileceği birçok alt problem bulunmaktadır. Bu problemlerden birisi de tedavinin olumlu sonuçlanabilmesi adına, hastanın sürekli uzman kontrolü altında olarak, verilen egzersiz setlerinin yeterli sayıda tekrar edilmesinin sağlanmasıdır. Bu probleme bilgisayar bilimleri açısından bakıldığında ise, insan eyleminin donanımlar vasıtasıyla doğru biçimde algılanması gerekmektedir. Bu tez çalışmasında, egzersiz kümelerinin kişilere daha yüksek motivasyon ile yaptırılabilmesi amacıyla, gerçek zamanlı olarak çalışabilen, insan eylemini doğru biçimde anlamlandırılabilen, çoklu derinlik sensörü kullanımı yaklaşımı barındıran, sanal gerçeklik ile birleştirilmiş hareket tanıma sistemi geliştirilmiştir. Geliştirilen sistemde gerçek dünyada bulunan kişinin, sanal dünyadaki ortam içerisinde temsil edilmesi sağlanarak daha etkili bir tedavi süreci için ortam oluşturulmuştur. Geliştirilen sistemde, eylemin başarılı biçimde tanımlanabilmesinin yanısıra, kişi tarafından yapılan eylem ile uzman kişiden elde edilen eylem arasındaki hatalar belirlenerek kişilere aktarılabilmektedir. Bu sayede egzersizin uygulanması anındaki hatalı öğrenmelerin de önüne geçilmesi hedeflenmektedir. Tez çalışması kapsamında, ayrıca çoklu derinlik sensörleri kullanılarak gerçekleştirilecek çalışmalar için, sensörlerin sahne üzerindeki yerleşim planlarını otomatik olarak hazırlayabilecek bir metot önerilmiştir. Çoklu derinlik sensörleri kullanılan çalışmalarda gözlemlenen deneysel ve öznel olarak oluşturulan sahne yerleşim planlarının yerine, belirli ölçütlere sahip sahne planının Yapay Arı Kolonisi yöntemi ile elde edilmesi sağlanmıştır.
