TOBB ETU GCRIS Database
Not a member yet
9778 research outputs found
Sort by
Knowledge Cohesion: Uniting Europe Through Research Networks
No full textResearch collaboration plays a pivotal role in not only disseminating existing knowledge but also catalyzing the generation of novel insights. This book delves into the benefits that arise from research collaborations, shedding light on their multifaceted impacts. It serves as a pioneering exploration into the nexus of collaboration, knowledge convergence, and knowledge cohesion, drawing extensively from the rich literature on the EU’s cohesion policy and collaboration-induced knowledge diffusion. Firstly, Knowledge Cohesion: Uniting Europe Through Research Networks unravels the nuanced interplay among collaboration, knowledge convergence, and knowledge cohesion. Secondly, it carves out clear distinctions between the realms of convergence and cohesion. Lastly, it unveils an empirical framework, offering tools for quantifying and analysing knowledge cohesion. These contributions culminate in the conceptualisation of knowledge cohesion, enriching our understanding of how collaborative research profoundly impacts the advancement of knowledge. The empirical analyses show that while the research collaboration network indicates knowledge convergence, there is no evidence for knowledge cohesion in Europe. This book stands as a resource for scholars, policymakers, and practitioners alike, inviting them to explore the nuanced interplay of research collaboration and knowledge dynamics, while presenting knowledge cohesion as a new concept. © 2024 İbrahim Semih Akçomak, Umut Yılmaz Çetinkaya, Erkan Erdil and Müge Özman
Bio-Inspired Thin-Walled Energy Absorber Adapted From the Xylem Structure for Enhanced Vehicle Safety
No full textThroughout evolution, plants and animals have optimized their structure to thrive in a wide range of extreme environments, offering natural structures with both low mass and high-energy absorption capacities. Vascular plants have developed a specialized tissue known as the Xylem, which offers structural support and facilitates the transport of water, mineral nutrients, and signals throughout the plant. This study aims to enhance the crashworthiness of vehicles by adapting the Xylem structure to design an effective bio-inspired thin-walled structure. Several different crash tube configurations are considered first, and their crashworthiness performances are assessed based on two different metrics: specific energy absorption (SEA)and crush force efficiency (CFE) , which are determined by using the finite element analysis software LS-DYNA. Then, the crash tube configuration with the best performance is chosen for further investigation. A surrogate-based optimization study is performed, and it is found that SEA and CFE are improved by 151% and 113% compared to an empty circular thin-walled crash tube. Furthermore, the simplified super folding element theory has been used for building a theoretical model that predicts the mean crushing force of the Xylem-mimicking structure. The simulation results and calculated values show a strong agreement, indicating that the proposed theoretical model is of high accuracy
Which Implant Is Better for the Fixation of Posterior Wall Acetabular Fractures: a Conventional Reconstruction Plate or a Brand-New Calcaneal Plate?
No full textBackground: Increased posterior wall acetabular fractures among older adults, require precise treatment to restore stability to the joint, lower the risk of degenerative arthritis, and enhance overall functional recovery. The purpose of this study was to compare the fixation stability and mechanical characteristics of calcaneal buttress plate and conventional reconstruction plate under different loading condition. Methods: Typical acetabular posterior wall fractures were created on twenty synthetic hemipelvis models. They were fixed with calcaneus plate and reconstruction plate. Dynamic and static tests were performed. Displacements of fracture line and stiffness were calculated. Findings: After dynamic loading, calcaneus plate fixation has significantly less displacement than the reconstruction plate on the superior posterior wall. Under static loading condition, the calcaneus plate group has significantly less displacement than the reconstruction plate group on the inferior posterior part of the fracture. The average stiffness values of the calcaneus plate group and the reconstruction plate group were 265.16 +/- 53.98 N/mm and 167.48 +/- 36.87 N/mm, respectively and a statistically significant difference was found between the two groups. Interpretation: The calcaneal plate group demonstrated better stability along the fracture line after dynamic and static loading conditions. Especially when the fragment was on the acetabulum's superior posterior, inferior posterior, and inferior rim, Calcaneal buttress plates offer biomechanically effective choices.Manisa Celal Bayar University [2020-076]We want to thank Manisa Celal Bayar University (project number 2020-076) for funding this research
Wideband Channel Capacity Maximization With Beyond Diagonal Ris Reflection Matrices
No full textFollowing the promising beamforming gains offered by reconfigurable intelligent surfaces (RISs), a new hardware architecture, known as beyond diagonal RIS (BD-RIS), has recently been proposed. This architecture enables controllable signal flows between the RIS elements, thereby providing greater design flexibility. However, the physics-imposed symmetry and orthogonality conditions on the non-diagonal reflection matrix make the design challenging. In this letter, we analyze how a BD-RIS can improve a wideband channel, starting from fundamental principles and deriving the capacity. Our analysis considers the effects of various channel taps and their frequency-domain characteristics. We introduce a new algorithm designed to optimize the configuration of the BD-RIS to maximize wideband capacity. The proposed algorithm has better performance than the benchmarks. A BD-RIS is beneficial compared to a conventional RIS in the absence of static path or when the Rician ;#x03BA;-factor is smaller than 5 dB. IEEEStiftelsen för Strategisk Forskning, SSF; Türkiye Bilimsel ve Teknolojik Araştırma Kurumu, TÜBİTAK, (FFL18-0277
Esnek Çinko Oksit-perovskit Temelli Fotodedektör Üretimi ve Karakteristik Özelliklerinin Analizi
No full textBu tez kapsamında, Mn katkılı ve katkısız ZnO-NÇ ile kararlı yapıda kurşunlu ve kurşunsuz inorganik halojenür perovskit tabanlı yarıiletkenlerin esnek alttaşlar üzerine depolanmasıyla katı-hal MSM fotodedektör ve yarı-katı olmak üzere kendinden güç üreten fotodedektör aygıtlar üretilmiştir. İlk olarak poliimit (PI) alttaş üzerine farklı molar konsantrasyonlarda (16,5 mM ve 3,5 mM) büyütülen ZnO-NÇ içerisine, molar hacimce %5 ve %9 olmak üzere Mn katkılanarak elde edilen ince filmlerin malzeme karakterizasyonları (XRD, PL, SEM ve UV-vis) incelenmiştir. Metal kontak olarak tasarlanmış interdijital Pt kontak ile belirlenen desenlerle kaplandıktan sonra ise fotodedektör aygıt olarak performasları incelenmiştir. Yapısal karakterizasyonlar sonucunda Mn atomlarının ZnO kristal kafes yapısına dahil olup performansı arttırdığı tespit edilmiştir. Fotodedektörlerin temel parametrelerin tespiti için, %5Mn:ZnO-NÇ3,5 fotodedektörün elektriksel ölçümlerinin alınmasıyla önemli performans parametreleri olan D*, R, S%, EQE(%), 15 V gerilim ve 367 nm dalga boyu ışık altında sırasıyla 6,38x1012, 10,768 A/W, 3x104 ve %3.645 olarak hesaplanmıştır. %5Mn:ZnO-NÇ3,5 fotodedektörün UV (367 nm/2,42 mW/cm2) ışık altında, yükselme zamanı (τr) ve alçalma zamanı (τd) değerleri sırasıyla 32,2 s ve 41,9 s olarak bulunmuştur. Beyaz ışık altında (840 mW/cm2) ise τr ve τd 0,3 s olarak hesaplanmıştır. Bu örneğe 9.500 defa içbükey, dış bükey bükme sonucunda, sağlamlık ve kararlılık sergilemiştir. %9Mn:ZnO-NÇ16,5 fotodedektörün D* değeri; 1,85x1014, R değeri; 159,467 A/W, S% değeri; 1,6x106 ve EQE(%) değeri; 54 olarak hesaplanmıştır. %9Mn:ZnO-NÇ16,5 fotodedektörün UV (367 nm) ışık altında, τr ve τd değerleri sırasıyla 21,1 s ve 57 s bulunmuştur. Beyaz ışık altında (840 mW/cm2) ise yük τr 0,91 s ve τd 4,19 s olarak hesaplanmıştır. Bu örneğe 10.000 defa içbükey, dış bükey bükme sonucunda, sağlamlık ve dayanıklılık göstermiştir. Esnek kendinden güç üreten yarı-katı-fotoelektrokimyasal (PEC) fotodedektör için İTOPET alttaş kullanılmıştır. İTOPET üzerine büyütülen Mn katkılı ZnO-NÇ inorganik halojenür perovskit olan CsPbBr3 tipi perovskit ile kaplanmış ve karbon kağıt ve İTOPET alttaşın karşıt elektrot olarak kullanılmasıyla sandviç formda birleştirilerek fotodedektör aygıt elde edilmiştir. İTOPET üzerine sentezlenen Mn katkılı ZnO-NÇ ve CsPbBr3 ilavesi ile elde edilen ince filmlerin UV-vis, SEM, XRD, EIS elementel haritalama ölçümleri alınmıştır. İTOPET/%5Mn:ZnO-NÇ/CsPbBr3 ince filmin 17 hafta UV-vis ölçümü alınmış, 538 nm dalga boyunda %7'lik küçük azalma olmuştur. %5Mn:ZnO-NÇ üzerine CsPbBr3 eklenmesi ile fotodedektör veriminde kayda değer artış gözlenmiştir. İTOPET/%5Mn:ZnO-NÇ/CsPbBr3 fotodedektörün, 400 nm (2,99 mW/cm2) dalga boyu ışık altında ve 0 volt D* değeri; 2,83x1012, R değeri; 95,12 mA/W, S% değeri; 8,04x106 ve harici kuantum verimi (EQE%); 29,55 olarak hesaplanmıştır. İTOPET/%5Mn:ZnO-NÇ/CsPbBr3 fotodedektörün UV (400 nm) ışık altında, yükselme zamanı ve alçalma zamanı değerleri, 10,62 ms ve 42,9 ms bulunmuştur. Bu fotodedektöre AM1,5 (100 mW/cm2) ışık altında 125 defa açma kapama ve 400 nm ışık altında 53 defa açma-kapama uygulanarak akım zaman ölçümleri alınmış ve kararlılık göstermiştir. Ayrıca farklı açılarda (35o, 45o ve 60o) 400 nm dalga boyu ışık altında 27 defa açma-kapama uygulanarak akım-zaman ölçümleri alınmıştır. AM1,5 ışk altında yükselme zamanı 13,33 ms ve alçalma zamanı 50,91 ms bulunmuştur. Ayrıca bu dedektörün 12 hafta sonunda yapılan akım-zaman ölçümünde hala fotodedektör özelliği sergilediği görülmüştür. Yarı-katı esnek kendinden güç üreten fotodedektör bir de çift katmanlı perovskit (Cs2AgBiBr6) kullanılarak üretilmiştir. ZnO-NÇ3K ve ZnO-NÇ3K Cs2AgBiBr6 ince filmlerin SEM analizi ve Cs2AgBiBr6 kristalin XRD ölçümü alınmıştır. İTOPET/ZnO-NÇ3K/Cs2AgBiBr6 fotodedektörün 0 V gerilim ve 382 nm dalga boyu ışık altında, yükselme zamanı 16,6 ms ve alçalma zamanı 8,76 ms bulunmuştur. Ayrıca farklı açılarda (35o, 45o ve 60o) 382 nm dalga boyu ışık altında 50 defa ışığı kapatıp açarak akım-zaman ölçümleri alınmıştır.Within the scope of this thesis, solid-state MSM photodetector and quasi-solid-state self-powered photodetector (PD) devices were produced on flexible substrates by depositing Mn-doped and pristine ZnO nanorods (ZnO-NR) and lead-based and lead-free all-inorganic halide perovskites. Flexible thin films were synthesized by doping 5% and 9% molar percent Mn into 16.5 mM and 3.5 mM ZnO on polyimide (PI) substrates, followed by depositing interdigital Pt contact designed as a metal contact. The synthesized flexible thin films based on Mn-doped and pristine ZnO-NR were characterized via XRD, PL, SEM, and UV-vis measurements to delve into the material properties. These structural characterizations indicated that Mn atoms were successfully inserted into the ZnO crystal lattice structure which resulted in enhancement in light absorption properties and the PD performance. The key parameters of the PD devices including D*, R, S%, and EQE(%) were evaluated after conducting electrical measurements. 5%Mn:ZnO- Upon 15 V of applied potential with 367 nm wavelength of light, NR3.5-based PD device D* has D*, R, S% and EQE(%) values as 6.38x1012, 10.768 A/W, S% 3x104 and 3,645, respectively. The rise time (τr) and fall time (τd) values of the 5%Mn:ZnO-NR3.5-based PD under UV (367 nm:2.42 mW/cm2) light were evaluated as 32.2 s and 41.9 s, respectively. Under white light illumination with 840 mW/cm2 of power density, the τr and τd values were determined as 0.3 s. In addition to this significant performance, the PD device exhibited durability and stability upon consecutive concave and convex bending for 9,500 times. D* value of 9%Mn:ZnO-NC16.5 photodetector; 1.85x1014, R value; 159.467 A/W, S% value; 1.6x106 and EQE(%) value; It was calculated as 54. The τr and τd values of the 9%Mn:ZnO-NR16.5 -PD at UV light illumination (367 nm) were found to be 21.1 s and 57 s, respectively. Under white light (840 mW/cm2), the τr and τd values were determined as 0.91 s and 4.19 s, respectively. This sample also showed remarkable durability upon concave and convex bending 10,000 times. İTOPET substrate was used for the flexible Quasi-solid-state PEC self-powered photodetector. After the synthesis of Mn-doped ZnO-NR on ITOPET, the CsPbBr3 perovskite structure was synthesized and a photodetector device was obtained. UV-vis, SEM, and UV-vis measurement of ITOPET/5%Mn:ZnO-NR/CsPbBr3 thin film was taken for 17 weeks, and there was a small decrease of 7% at the 538 nm wavelength. A significant increase in PD efficiency was observed with the addition of CsPbBr3 onto 5%Mn:ZnO-NR. D* value of ITOPET/5%Mn:ZnO-NR/CsPbBr3 photodetector under 400 nm (2.99 mW/cm2) wavelength light with no applied bias, D*, R, S% and EQE(%) values were evaluated as 2.83x1012, 95.12 mA/W, 8.04x106 and 29.55, respectively. The rise time and fall time values of the ITOPET/5%Mn:ZnO-NC/CsPbBr3 photodetector under UV (400 nm) light were found to be 10.62 ms and 42.9 ms. This photodetector was turned on and off 125 times under AM1.5 (100 mW/cm2) light and on and off 53 times under 400 nm light, and current-time measurements were taken, and stability was demonstrated. In addition, current-time measurements were taken by turning on and off 27 times under 400 nm wavelength light at different angles (35o, 45o, and 60o). Under AM1.5 light, the rise time was 13.33 ms and the fall time was 50.91 ms. In addition, it was observed that this detector still exhibited photodetector properties in the current-time measurement made at the end of 12 weeks. The quasi-solid-state flexible and self-powered PD device was produced using a double-layer perovskite (Cs2AgBiBr6). SEM analysis of ZnO-NR3K and ZnO-NR3K Cs2AgBiBr6 thin films and XRD measurement of Cs2AgBiBr6 crystal were taken. The rise time of the ITOPET/ZnO-NR/Cs2AgBiBr6 photodetector under 0 V voltage and 382 nm wavelength light was found to be 16.6 ms and the fall time was 8.76 ms. In addition, current-time measurements were taken by turning the light on and off 50 times under 382 nm wavelength light at different angles (35o, 45o and 60o)
Interest, Need, or Reputation? Determinants of Qatar's Foreign Aid
No full textThe literature on the motivations behind emerging donors' foreign aid contributions-with the exception of larger countries such as China, India, Brazil and Turkey-lacks original data and empirical analyses. This article addresses this gap by providing a novel, detailed dataset of the foreign aid allocations of one resilient, relatively small emerging donor, Qatar, for the period from 2014 to 2021. The contents of our database, which we dub Qatar Aid Database, encapsulate the features of Qatar's international aid. To illustrate our dataset's efficacy, we investigate whether Qatar's foreign aid operations aim to satisfy donor interests, recipient needs, or something else. Our empirical findings affirm that although Qatar's aid allocation decisions prioritize recipient needs, its aid provision depends more on Qatar's foreign policy interests in the targeted countries. ; Agrave; l'exception des plus gros pays tels que la Chine, l'Inde, le Br ; eacute;sil et la Turquie, la litt ; eacute;rature relative aux motivations qui se cachent derri ; egrave;re les contributions d'aide ; eacute;trang ; egrave;re des donateurs ; eacute;mergents manque de donn ; eacute;es in ; eacute;dites et d'analyses empiriques. Cet article rem ; eacute;die ; agrave; cette lacune en fournissant un ensemble de donn ; eacute;es in ; eacute;dit et d ; eacute;taill ; eacute; sur les attributions d'aides ; eacute;trang ; egrave;res de 2014 ; agrave; 2021 d'un donateur ; eacute;mergent r ; eacute;silient et relativement petit : le Qatar. Le contenu de notre base de donn ; eacute;es, que nous appelons QATARAID, regroupe les caract ; eacute;ristiques de l'aide internationale qatarienne. Pour illustrer l'efficacit ; eacute; de notre ensemble de donn ; eacute;es, nous nous int ; eacute;ressons ; agrave; l'objectif des op ; eacute;rations d'aide ; eacute;trang ; egrave;re du Qatar : satisfaire les int ; eacute;r ; ecirc;ts des donateurs, les besoins des b ; eacute;n ; eacute;ficiaires ou autre chose ? Nos r ; eacute;sultats empiriques affirment que bien que les d ; eacute;cisions d'attributions d'aide du Qatar accordent la priorit ; eacute; aux besoins des b ; eacute;n ; eacute;ficiaires, sa fourniture d ; eacute;pend plus des int ; eacute;r ; ecirc;ts de politique ; eacute;trang ; egrave;re qatariens dans les pays cibles. La literatura en materia de las motivaciones que se encuentran detr ; aacute;s de las contribuciones de ayuda exterior por parte de pa ; iacute;ses donantes emergentes (con la excepci ; oacute;n de pa ; iacute;ses m ; aacute;s grandes como China, India, Brasil y Turqu ; iacute;a) carece de datos originales y de an ; aacute;lisis emp ; iacute;ricos. Este art ; iacute;culo aborda esta brecha debido a que proporciona un conjunto de datos novedoso y detallado de las asignaciones de ayuda exterior por parte de un donante emergente, resistente y relativamente peque ; ntilde;o, Catar, para el per ; iacute;odo entre 2014 y 2021. El contenido de nuestra base de datos, a la que denominamos QATARAID, resume las caracter ; iacute;sticas de la ayuda internacional otorgada por Catar. Con el fin de ilustrar la eficacia de nuestro conjunto de datos, investigamos si las operaciones de ayuda exterior que lleva a cabo Catar tienen como objetivo satisfacer sus intereses como donante, las necesidades de los receptores o alg ; uacute;n otro motivo diferente. Nuestros hallazgos emp ; iacute;ricos afirman que, si bien las decisiones de asignaci ; oacute;n de ayuda por parte de Catar priorizan las necesidades de los receptores, su provisi ; oacute;n de ayuda depende m ; aacute;s de los propios intereses de Catar en materia de pol ; iacute;tica exterior en los pa ; iacute;ses objetivo.The authors express sincere gratitude to Nihat Muurtay and Orcun Demir for their invaluable research assistance on this article. We also extend our appreciation to Saban Karda and ozgur ozdamar for their insightful feedback on earlier drafts. Special thanks are due to the editors of Foreign Policy Analysis and the anonymous peer reviewers for their invaluable feedback and constructive suggestions
Elektrikli Otobüslerde Çizge Tabanlı Öznitelik Seçimi ve Makine Öğrenmesi ile Kestirimci Bakım
No full textThis thesis employs a graph-based feature selection method to analyse the relationships between the characteristics of a CAN (Controller Area Network) Bus data set pertaining to electric buses. Furthermore, it investigates the prediction performance of targeted alarms utilising artificial intelligence techniques. The study employed data obtained from CAN Bus systems of diverse vehicles over an extended period. The raw data packages underwent extensive pre-processing techniques to create data sets suitable for the structural needs of machine learning models. In this process, a hybrid graph-based feature selection tool was developed using a combination of statistical filtering methods, including Pearson correlation analysis, Cramer's V statistical table and ANOVA F-test, and optimisation-based community detection algorithms, such as InfoMap, Leiden, Louvain and Fast Greedy. The developed tool was employed to identify feature subsets, evaluate the relational position of the features and their relationship with the targeted alarms, select distinctive features and perform dimension reduction. In the selection of machine learning models, a number of different approaches were considered, including neural network architectures and traditional classifiers. Amongst these, the Support Vector Machine (SVM), Random Forest (RF), eXtreme Gradient Boosting (XGBoost), Long Short-Term Memory (LSTM), and Recurrent Neural Network (RNN) models were applied, and their performance was then evaluated. In order to mitigate the impact of class imbalance on model performance, SMOTEEN (Synthetic Minority Oversampling Technique-Edited Nearest Neighbours) and binary search-based time interval down-sampling methods were employed. In order to identify the optimal distinctive classifier space for each model, a grid search and random search-based hyperparameter optimisation (hyperparameter tuning) was conducted. The performance of the developed models was evaluated based on the metrics employed for model validation. The predictions of the models were explained by examining the weights (importance) of the features and utilising the LIME (Local Interpretable Model-Agnostic Explanations) tool to identify the features that triggered the predictions, thereby gaining insights into the functionality (design) of the system. The present study has investigated the prediction performance of specific alarms in vehicles utilising CAN Bus data with machine learning techniques. Additionally, a graph-based feature selection tool has been developed, enabling the visual interpretation of relationships between features and facilitating the enhancement of corrective, preventive, predictive and proactive maintenance (design improvement) approaches. It is concluded that the techniques employed throughout the machine learning life cycle, as conducted in the study, can provide valuable insights and contribute to the advancement of maintenance strategies through machine learning, particularly in contexts where data quality and access to expert opinion may present challenges, in alignment with the principles of Industry 4.0. Keywords: Predictive maintenance (PdM), Machine learning (ML), Graph based feature selection, Explainable artificial intelligence (XAI).Bu tez çalışmasında, elektrikli otobüslere ait CAN (Controller Area Network) Bus verileri kullanılarak çizge tabanlı öznitelik seçim yöntemi ile sistemin özellikleri arasındaki ilişkiler analiz edilmiş ve üretici tarafından hedeflenen alarmların yapay zekâ teknikleri ile kestirim performansı araştırılmıştır. Araştırmada belli bir süre boyunca, farklı araçlardan elde edilen CAN Bus verileri kullanılmıştır. Ham veri paketleri üzerinde kapsamlı ön işleme teknikleri uygulanmış ve makine öğrenmesi modellerinin yapısal ihtiyaçlarına uygun veri setleri oluşturulmaya gayret edilmiştir. Bu süreçte Pearson korelasyon analizi, Cramer's V istatistik tablosu ve Anova F-testi gibi istatiksel filtreleme yöntemleri ve InfoMap, Leiden, Louvain, Fast Greedy gibi optimizasyon tabanlı topluluk tespiti algoritmaları kullanılarak, hibrit çizge tabanlı öznitelik seçim aracı geliştirilmiştir. Geliştirilen araç kullanılarak özellik altkümeleri belirlenmiş, özelliklerin ilişkisel konumu ve hedeflenen alarmlarla ilişkileri değerlendirilerek ayırt edici öznitelikler seçilerek boyut azaltımı yapılmıştır. Makine öğrenmesi modellerinin seçilmesinde, Destek Vektör Makinesi (SVM), Rastgele Orman (RF), eXtreme Gradient Boosting (XGBoost), Uzun Kısa Süreli Bellek (LSTM), ve Tekrarlayan Sinir Ağları (RNN) modelleri dahil olmak üzere sinir ağı mimarileri ve geleneksel sınıflandırıcılar uygulanarak performansları değerlendirilmiştir. Sınıf dengesizliğinin modellerin performansındaki etkisini minimize etmek için SMOTEEN (Synthetic Minority Oversampling Technique-Edited Nearest Neighbours) ve ikili arama tabanlı zaman aralığı örnekleme (binary search based time interval down-sampling) yöntemleri uygulanmıştır. Her modelin en iyi ayırt edici sınıflandırıcı uzayını bulmak için Izgara Arama (Grid Search) ve Rastgele Arama (Random Search) tabanlı hiperparametre optimizasyonu (Hyperparameter tuning) yapılmıştır. Geliştirilen modellerin performansları geçerleme (validation) metrikleri esas alınarak değerlendirilmiştir. Modellerin kestirimleri özniteliklerin ağırlıkları incelenerek ve LIME (Local Interpretable Model-Agnostic Explanations) aracı kullanılarak açıklanmaya çalışılmış, kestirimleri tetikleyen öznitelikler belirlenmiş ve sistemin işleyişi (tasarımı) ile ilgili içgörüler elde edilmiştir. Bu çalışma kapsamında CAN Bus verileri kullanılarak araçlardaki belli alarmların makine öğrenmesi teknikleri ile kestirim performansı araştırılmış, çizge tabanlı bir öznitelik seçme aracı geliştirilerek düzeltici, önleyici, kestirimci ve proaktif bakım (tasarımın iyileştirilmesi) yaklaşımlarının iyileştirilmesine yardımcı olacak şekilde öznitelikler arasındaki ilişkilerin görsel olarak yorumlanması imkânı sağlanmıştır. Çalışmada izlenen makine öğrenmesi yaşam döngüsü kapsamında uygulanan tekniklerin farklı sektörlerde ve Endüstri 4.0 gereği bakım stratejilerinin makine öğrenimi yoluyla iyileştirilmesinde, özellikle veri kalitesi ve uzman görüşüne erişim zorluğu olan projelerde yararlı olabileceği değerlendirilmektedir. Anahtar Kelimeler: Kestirimci bakım (PdM), Makine öğrenimi (ML), Çizge tabanlı öznitelik seçimi, Açıklanabilir yapay zekâ (XAI)
Recent Progress in Aluminium Matrix Composites: a Review on Tribological Performance
No full textAluminium (Al) matrix composites are widely used in the automotive and aerospace sectors due to their low density, high specific strength, and excellent tribological performance. Although wear studies on Al matrix composites started approximately 40 years ago, they still continue to attract. This review reports the latest developments on the wear performance of up-to-date Al matrix composites. For this reason, the effect of different factors such as applied load, sliding speed, sliding distance, temperature, particle content, and particle size on tribological performance was investigated, considering the coefficient of friction and wear rate. Dominant wear mechanisms depending on wear parameters were also evaluated. Consequently, the review provides the challenges and future research directions in the tribology of Al matrix composites. © The Indian Institute of Metals - IIM 2024
Detection of Alzheimer’s and Parkinson’s Diseases Using Deep Learning-Based Various Transformers Models †
No full textAlzheimer’s disease is a neurodegenerative condition primarily attributed to environmental factors, abnormal protein deposits, immune system dysregulation, and the consequential death of nerve cells in the brain. On the other hand, Parkinson’s disease manifests as a neurological disorder featuring primary motor, secondary motor, and non-motor symptoms, accompanied by the rapid demise of cells in the brain’s dopamine-producing region. Utilizing brain images for accurate diagnosis and treatment is integral to addressing both conditions. This study harnessed the power of artificial intelligence for classification processes, employing state-of-the-art transformer models such as Swin transformer, vision transformer (ViT), and bidirectional encoder representation from image transformers (BEiT). The investigation utilized an open-source dataset comprising 450 images, evenly distributed among healthy, Alzheimer’s, and Parkinson’s classes. The dataset was meticulously divided, with 80% allocated to the training set (390 images) and 20% to the validation set (90 images). Impressively, the classification accuracy surpassed 80%, showcasing the efficacy of transformer-based models in disease detection. Looking ahead, this study recommends delving into hybrid and ensemble models and leveraging the strengths of multiple transformer-based deep learning architectures. Beyond contributing crucial insights at the intersection of artificial intelligence and neurology, this research emphasizes the transformative potential of advanced models for enhancing diagnostic precision and treatment strategies in Alzheimer’s and Parkinson’s diseases. It signifies a significant step towards integrating cutting-edge technology into mainstream medical practices for improved patient outcomes. © 2024 by the author