Mühendislik Fakültesi Koleksiyonu

Bu koleksiyon için kalıcı URI

https://hdl.handle.net/11501/18

Listeleniyor 1 - 20 / 427

A novel hyperbolic tangent-augmented controller framework for temperature control in jacketed continuous stirred tank reactors
(Nature Research, 2026) İzci, Davut; Ekinci, Serdar; Ökten, İrfan; Çınar, Rıdvan Fırat; Rashdan, Mostafa; Salman, Mohammad; Bektaş Güneş, Burcu; Ahmad, Mohd Ashraf
Accurate temperature regulation of jacketed continuous stirred tank reactors (CSTRs) remains a challenging task due to strong nonlinearities, tight coupling between mass and energy balances, and sensitivity to disturbances and operating-point variations. In this study, a novel augmented proportional–integral–derivative (PID) controller incorporating a hyperbolic tangent nonlinearity (APID-T) is proposed for robust temperature control of an exothermic CSTR. The controller structure extends the classical PID framework by embedding a bounded nonlinear term that enhances transient shaping and robustness while preserving simplicity and practical implementability. The tuning of the APID-T parameters is formulated as a constrained nonlinear optimization problem, where a composite objective function combining normalized overshoot and integral squared error is minimized. To solve this problem efficiently, the recently developed Schrödinger optimizer (SRA) is employed, exploiting its balanced exploration–exploitation mechanism. A detailed nonlinear dynamic model of the jacketed CSTR is considered, and stability characteristics around the nominal operating point are examined to ensure meaningful closed-loop operation. The proposed SRA-based APID-T design is extensively evaluated through comparative simulations against several state-of-the-art metaheuristic optimizers and alternative controller structures, including PI, PID with filter, two-degree-of-freedom PID, and fractional-order PID controllers. Performance is assessed using statistical indicators, convergence behavior, and time-domain response metrics under identical optimization settings. In addition, widely used error performance criteria, including the integral squared error, integral time absolute error, and integral time squared error, are computed to provide a comprehensive quantitative assessment of the tracking performance. The results demonstrate that the SRA-tuned APID-T controller consistently achieves lower objective-function values, faster convergence, reduced settling time, and significantly smaller overshoot compared with the competing approaches. Furthermore, frequency-domain analysis based on the Bode characteristics of the linearized open-loop system confirms favorable stability margins, supporting the robustness of the proposed control structure. Additional stability and robustness evaluations are conducted under practical non-ideal conditions, including feed-temperature disturbances, measurement noise, and multiple setpoint variations, where the controller maintains stable and accurate temperature regulation across the considered operating scenarios.
Corrosion studies of additive manufactured lattice architectures
(Springer Nature, 2026) Gürkan, Doruk; Sağbaş, Binnur
The research provides an analysis of the corrosion of additive manufactured metallic lattice structures. Additive manufacturing provides an opportunity to generate complex lattice geometries such as strut based, surface based and shell based. These types of lattice structures provide immense potential for applications from implants to biomedical devices and components of aircraft. The paper delves into details of factors like corrosion-resistant properties of thin protective film on metallic surfaces and localized corrosion mechanisms of pits. In addition to these factors, the role of galvanic corrosion in various domains of metallic lattice structures has been discussed. The role of lattice structure design on corrosion properties has been discussed. Geometry parameters like relative density and cell size determine electrolyte flow patterns. The influence of strut and cell size on ion movement in electrolytes makes it evident that there is an important balance to be achieved for an additive manufacturing component to possess high strength as well as high corrosion-resistant properties. Additional processing and application of coatings on surfaces of metallic components improve corrosion-resistant properties. These techniques improve the strength of the biofilm on metallic surfaces by eliminating surface irregularities that act as primary corrosion site domains.
Performance of Laser Powder Bed Fusion (LPBF)-based coolers for semi-conductive thermoelectric modules
(Springer Nature, 2026) Dilibal, Savaş; Demiröz, Ömer Buğra
The performance and longevity of semi-conductive thermoelectric modules depend heavily on efficient thermal management, which is essential across applications in electronics, battery systems, and biomedical technologies. Conventional manufacturing approaches for compact cooling structures are limited by geometric complexity, machining difficulties, and high prototyping costs. In this study, Laser Powder Bed Fusion (LPBF)-based additively manufactured passive AlSi10Mg coolers were designed, fabricated, and thermally evaluated. Three distinct geometries with equal fill ratios were designed to optimize heat dissipation across thermoelectric module surfaces. Numerical simulations verified thermal distribution prior to fabrication, ensuring alignment between design parameters and module requirements. The LPBF-based additive manufacturing enabled the precise realization of complex geometries, producing components with high dimensional accuracy, excellent surface quality, and thermal conductivity suitable for integration with TEC-12710 modules. Following fabrication, the cooling structures were mounted to the thermoelectric modules with thermal paste and tested under controlled conditions. Results confirmed that LPBF-fabricated geometries enhanced heat transfer efficiency, accelerated surface cooling, and provided more uniform temperature distribution compared to conventional designs. The findings demonstrate the potential of LPBF as a versatile and cost-effective solution for manufacturing tailored thermal management coolers for thermoelectric components, with promising applications in mechatronic devices, medical electronics, energy storage, and compact thermal regulation systems.
Application of hybrid meta-heuristic-based data-driven models to forecast streamflow drought index
(Springer, 2026) Ghasemlounia, Redvan; Gharehbaghi, Amin; Ahmadi, Farshad
Drought is a complex natural catastrophe threatened markedly the societies. Thus, its precise prediction has a noteworthy effect in numerous sections such as water resources, nutrition, economics, industry, etc. In the current investigation, to evaluate the hydrological drought procedure in Nazlu River basin, Urmia City, West Azerbaijan province, the streamflow drought index (SDINRB) is applied. In this direction, the SDINRB in four different time measures including 3, 6, 9, and 12-month are calculated using 829 mean monthly streamflow datasets recorded from Aug 1951 to Aug 2020 by Tapik hydro-meteorological station. Then, two robust advanced hybrids support vector regression (SVR) with Harris hawk algorithm (HHO) and intelligent water drop (IWD) optimization algorithms i.e., hybrid SVR-HHO and SVR-IWD models, are developed to estimate the fluctuations pattern of SDINRB-3-12. Given this, to accomplish the optimum supportive models' structure, many scenarios are implemented by tuning meta-parameters such as a number of hidden neurons and deterministic factors of SVR, HHO, IWD algorithms. According to the performance assessment criteria and comparison plots, the hybrid SVR-RBF-HHO model under ideal meta-parameters is identified as the suitable model for predicting SDINRB-3-6 droughts, yet the hybrid SVR-RBF-IWD model is recognized as the appropriate model for forecasting SDINRB-9-12 droughts. Likewise, the best modelling performance is achieved by the hybrid SVR-RBF-HHO model in predicting SDINRB-6 drought. It results in an RMSE, R-2, NSE, and MBE of 0.23, 0.94, 0.92, and 0.072, respectively. Nonetheless, for the single SVR-RBF as the benchmark model is attained 0.43, 0.79, 0.78, and 0.045, respectively.
Yapay zekâ okuryazarlığının kişisel başarı ve kariyer kararlılığına etkisinin araştırılması: lise öğrencileri örneği
(Akademik Bilişim Vakfı, 2025) Özer, Tuğba; Suvay Eker, Halime
Yapay zekâ, günümüzde hızla gelişmekte olan ve pek çok alanda etkisini gösteren önemli bir teknolojidir. Eğitim, sağlık ve kurumsal sektör gibi çeşitli alanlarda köklü değişiklikler gerçekleştiren yapay zekâ, bireylerin kişisel başarı ve kariyer tercihleri üzerinde de belirleyici bir rol oynayabilmektedir. Bu araştırmada lise öğrencilerinin yapay zekâ okuryazarlık düzeylerinin kişisel başarı ve kariyer kararlılığı üzerindeki etkisini belirlemek amaçlanmış, bununla birlikte yapay zekâ okuryazarlığının demografik değişkenlere göre farklılık gösterip göstermediği araştırılmıştır. Dört bölümden oluşan anket aracılığı ile Antalya ilindeki 472 lise öğrencisinden toplanan veriler sırasıyla doğrulayıcı faktör analizi, yapısal eşitlik modellemesi ve parametrik olmayan testler ile analiz edilmiştir. Alınan sonuçlarda, yapay zekâ okuryazarlığının hem kişisel başarı hem de kariyer kararlılığı ile anlamlı ve pozitif ilişkiler içinde olduğu görülmüştür. Ayrıca, yapay zekâ okuryazarlığı cinsiyet, yaş ve lise türüne göre farklılık göstermezken, sınıf düzeyi ve internet kullanım sıklığına göre anlamlı farklılıklar göstermiştir.
Endüstriyel makinelerin arıza durumlarına göre segmentasyonu: K-means ve fuzzy C-means algoritmaları ile RFM analizi
(Gazi Üniversitesi, 2025) Canlı, Hikmet; Varıcı, Sena
Bu çalışma makinelerin segmentasyonunu, bakım ve arıza kayıtlarına dayalı olarak RFM analizi ile değerlendirdikten sonra K-means ve Fuzzy C-means kümeleme algoritmaları kullanarak değerlendirmeyi amaçlamaktadır. Her bir makinenin arıza geçmişi makinelerin bakım ve arıza verileri analiz edilerek incelenmiştir. Makinelerin segmentasyonunu değerlendirmek amacıyla Arıza Frekansı, Toplam Arıza Süresi ve Son Arıza zamanı gibi parametreler kullanılmıştır. Bu parametreler, müdahale edilmesi gereken makinelerin belirlenmesini ve makinelerin operasyonel sağlık durumlarını anlaşılmasını sağlamıştır. Makine verileri üzerinde RFM analizi uygulandıktan sonra K-means ve Fuzzy C-means algoritmaları kullanılarak kümeleme yapılmıştır. Bu çalışma, makinelerin bakım süreçlerini optimize etmek, arıza eğilimlerini daha doğru tahmin etmek, operasyonel verimliliği artırmak ve maliyetleri düşürmek için veri odaklı bir yaklaşım sunmaktadır. Çalışma sonuçları David-Bouldin Index, Dunn Index, Calinski- Harabasz Index gibi metrikler kullanılarak kıyaslanmış ve en iyi kümelemeyi yapan algoritma seçilmiştir. Sonuçlar, makinelerin segmentlere ayrılmasını ve her segment için özel bakım ve iyileştirme stratejilerinin geliştirilmesini sağlamaktadır.
Assessment of water quality in Kirkuk city: using geographic information systems and multi-linear regression analysis
(MIM Research Group, 2026) Noori, Shaho Kh.; Naser, Ibrahim J.; Ghasemlounia, Redvan; Ibrahim, Mohammed O.; Raheem, Aram M.; Fayyadh, Imran A.
Climate change, pollution, and the degradation of water quality are making it harder to provide safe drinking water globally. For ensuring water safety, it is important to analyze its chemical and physical properties. Assessing the suitability of tap water for consumption according to World Health Organization (WHO) standards and Iraqi standards using GIS techniques and multi-linear regression analysis in Kirkuk City, Iraq, the study examines fourteen key water parameters across multiple city locations and conducted over 2020 and 2022, including pH, turbidity, conductivity, alkalinity, hardness, various ions, dissolved solids, temperature, and chlorine. GIS mapping was used to visualize the spatial distribution of these parameters, while multi-linear regression analysis was employed to identify statistical relationships between turbidity and other water quality indicators. The pH level of tap water in Kirkuk has been shown to be slightly to moderately alkaline, and turbidity exceeded acceptable limits in some locations and decreased slightly in 2022 from that in 2020. It was observed that total dissolved solids decreased during the period of this study and were under permissible standard limits. Electrical conductivity and total hardness remained stable and acceptable. The concentrations of most of the ions, including calcium, chloride, magnesium, sulfate, sodium, and chlorine, were lower than those of WHO and Iraqi standards. Potassium levels varied, meeting standards in some areas in 2020 but falling below WHO recommendations everywhere by 2022. The regression analysis revealed that turbidity was significantly influenced by total dissolved solids, electrical conductivity, and total hardness, with a strong correlation (R2 = 0.939), indicating that changes in water quality parameters are interdependent. However, the R2 value for single linear regression suggests potential variability, emphasizing the need for further data refinement and additional explanatory variables. The study highlights the need for ongoing monitoring and improvement of Kirkuk's water treatment processes, demonstrating the value of GIS-based spatial analysis combined with statistical modeling for identifying areas with compromised water quality and implementing targeted interventions to address these issues.
A comparative study on novel hybrid approaches based on CEEMDAN, random forest, deep learning methods for predicting daily wind speed
(Springer Science and Business Media Deutschland GmbH, 2026) Gharehbaghi, Amin; Ghasemlounia, Redvan; Ahmadi, Farshad; Mirabbasi, Rasoul; Haghighi, Ali Torabi
In this study, different kinds of hybrid Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) algorithms with forecasting models including Random Forest (RF), Gated Recurrent Unit (GRU), and Long Short-Term Memory (LSTM) neural networks, are developed to estimate the mean daily wind speed at the height of 2 m in A & gbreve;r & imath; city (WSst12), Turkey. In these hybrid models, different layer networks of single and integrated LSTM and GRU models include general single LSTM, general single GRU, simple coupled LSTM-GRU, and novel coupled LSTM with GRU through Addition layer (i.e., LSTM + GRU model) structures are applied. The most effective parameters on the WSst12, from a list of on-site potential meteorological parameters and wind speed values in its adjacent cities of A & gbreve;r & imath; province from Jan 2015-Dec 2019 through the Pearson correlation coefficient method, are determined. In the hybrid CEEMDAN and DNNs-based models, State activation functions (SAF), numbers of hidden neurons (NHN), dropout rates (P-rate), and network structural architect (NSA) as the meta-parameters are tuned for lessening the impact of overfitting/underfitting dilemmas and improving modeling performance. According to the comparison plots, performance evaluation measures, and total learnable parameter (TLP), the novel developed hybrid CEEMDAN-RF-(LSTM + GRU) model is confirmed as the best approach with an R2 of 0.86 while, in the optimal scenario using the RF model, R2 was 0.47.Graphical AbstractBased on the graphical snapshot, this study focuses on estimating daily mean wind speed at a 2-meter height in A & gbreve;r & imath;, Turkey, using hybrid data-driven models. The research integrates the Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) algorithm with advanced forecasting techniques, including Random Forest (RF), Long Short-Term Memory (LSTM), and Gated Recurrent Unit (GRU) neural networks. The modeling framework explores various configurations, such as standalone LSTM and GRU, coupled LSTM-GRU structures, and a novel LSTM + GRU model using an Addition layer to enhance predictive accuracy.
Advancement in forecasting rainfall-runoff process: application of a novel hybrid FMD-SVR-ABC modeling technique
(Springer International Publishing AG, 2026) Ghasemlounia, Redvan; Gharehbaghi, Amin; Ahmadi, Farshad
The prediction of monthly rainfall-runoff time series has a significant influence in planning and developing water resources projects. Thus, in this research, a novel advanced coupled predictive disintegration-optimization-based model is developed to improve the forecasting exactness of the mean monthly river's runoff. The suggested estimation model is a coupled version of the feature mode decomposition (FMD) algorithm and support vector regression (SVR) model optimized with artificial bee colony (ABC) metaheuristic algorithm, i.e., hybrid FMD-SVR-ABC model. Its performance is tested on monthly Barandouzchay River's runoff (BCRRm) watershed in Urmia City, West Azerbaijan Province from Sep 1971 to Aug 2022. In the FMD-based approaches, the optimal amount of mode number for the rainfall time series measured is 5. Using the partial autocorrelation function (PACF) technique, the number of predictor variables is determined as 9. Comparison plots and performance assessment criteria attest that the recommended model under the optimum predictor and meta-parameters tuned, provides better forecasting results with coefficient of determination (R2) of 0.82, root mean square error (RMSE) of 2.67 (m3/s), mean bias error (MBE) of 0.22 (m3/s), Nash-Sutcliffe efficiency (NSE) of 0.8. Comparatively, the individual SVR model leads to the R2 of 0.36, RMSE of 5.39 (m3/s), MBE of 2.23 (m3/s), and NSE of 0.23. Integrating with FMD and ABC algorithms lessens the RMSE value in the single SVR (as the benchmark model) by 27.8% and 15.7%, respectively. Therefore, the suggested hybrid model can be operated as an ingenious, sensible, and precise predictive model for the evaluation of the sequential rainfall-runoff rivers data, mainly the peak flows in different hydro-climatic regimes.
Reducing defect rates in smart manufacturing processes: deep learning-based quality control for automotive tow hooks
(PeerJ Inc., 2026) Canlı, Hikmet; Varıcı, Sena
In the automotive industry, quality control in the production of critical components is of great importance. One such critical component is the tow hook, a specialized connecting element mounted on the rear of motor vehicles, enabling them to safely tow another vehicle or load. Defective tow hooks can lead to both safety hazards and economic losses. This study aims to compare the performance of deep learning-based models for the classification of defective parts in industrial production processes. The constructed dataset includes defects such as burrs, dents on the thread, non-filling and dents on the surface. Additionally, a class for intact parts was created, resulting in a total of five categories. These categories were used to train and comprehensively evaluate models with different architectures, including residual neural network (ResNet), Mobile Convolutional Neural Network (MobileNet), Visual Geometry Group (VGG), and conventional convolutional neural networks (CNNs), using validation metrics. Experimental results indicate that the ResNet34 model achieved the highest performance, with a 100% accuracy rate and precision, recall, and F1-score values of 1.00 across all classes. The MobileNetV2 model achieved 96% accuracy, with class-wise validation metrics exceeding 0.90, demonstrating strong performance. These findings suggest that the proposed models can effectively distinguish between defective and intact parts with high accuracy, offering a reliable solution for industrial quality control applications. Future work will focus on enhancing the models' generalization capabilities through the evaluation of larger datasets and diverse production conditions.
Electrical and dielectric tailoring of glass fiber-reinforced concrete using ZnO-based hybrid nanocomposites
(Springer, 2026) Ramazanoğlu, Doğu; Musatat, Ahmad Badreddin; Subaşı, Azime; Demir, Ahmet; Subaşı, Serkan; Maraşlı, Muhammed
This study investigates frequency-dependent dielectric and electrical transport properties of glass fiber-reinforced concrete (GFRC) systematically doped with ZnO-based hybrid composite (ZnO-@) nanoparticles at 1%, 2%, and 3% mass fractions. Electrical impedance spectroscopy (20 Hz-5 MHz) coupled with microstructural characterization (SEM-EDX, FTIR) and mechanical validation establishes concentration-dependent polarization mechanisms governing electromagnetic property modulation. The 2% ZnO-@ formulation exhibits optimal dielectric enhancement with maximum real permittivity (epsilon '), superior AC conductivity (100 Hz-10 kHz domain), and 100% imaginary modulus augmentation (M ''), attributed to Maxwell-Wagner-Sillars interfacial polarization at ZnO-cement matrix boundaries. Equivalent circuit modeling reveals that grain boundary resistance escalates to 5.8 M Omega at optimal doping, and constant phase element (CPE) exponent values (P = 0.77-0.84) confirming non-Debye relaxation due to hierarchical microstructural heterogeneity. The critical percolation threshold, between 2% and 3% ZnO concentration, demarcates the transition from capacitive to conductive behavior, where specimens at 3% exhibit dielectric parameter regression toward baseline values due to nanoparticle agglomeration and the formation of conductive pathways. Spectroscopic validation confirms the integration of wurtzite-phase ZnO (Zn-O: 474 cm(-)1) with preserved calcium silicate hydrate phases, while post-aging Leeb hardness measurements demonstrate 171-176% mechanical reinforcement (387-456 HLD), validating the retention of structural durability. These findings establish quantitative compositional guidelines for engineering multifunctional construction composites with tailored electromagnetic response characteristics for interference shielding, capacitive energy storage, and electromagnetically compatible innovative infrastructure applications.
Reliable parameter estimation of nonlinear chaotic systems and PMSMs with the stellar oscillation optimizer
(Nature Research, 2026) Ekinci, Serdar; İzci, Davut; Jabari, Mostafa; Elsayed, Fahmi; Salman, Mohammad; Bektaş Güneş, Burcu
Accurate parameter identification is a critical prerequisite for reliable modeling, analysis, and control of nonlinear dynamical systems. This study introduces the stellar oscillation optimizer (SOO), a recently proposed metaheuristic inspired by the oscillatory behavior of stars, and investigates its effectiveness in estimating system parameters through a unified optimization framework. The identification problem is formulated as the minimization of a trajectory-mismatch cost function, where candidate solutions are iteratively refined by the oscillatory dynamics of SOO. To comprehensively evaluate its performance, four benchmark systems were considered: three canonical chaotic models (Lorenz, Chen, and R & ouml;ssler) and a practical engineering case represented by a permanent-magnet synchronous motor (PMSM). The outcomes were benchmarked against several state-of-the-art algorithms, including Kirchhoff's law algorithm (KLA), Tianji's horse racing optimization (THRO), puma optimizer (PO), and hiking optimization algorithm (HOA), under a standardized protocol. The results show that SOO consistently achieves numerically convergent solutions with machine-precision-level residuals under deterministic and noise-free simulation settings, while maintaining strong robustness across independent runs. In chaotic benchmarks, the reported residuals approach floating-point limits, which indicates stable numerical convergence rather than guaranteed physical identifiability under real measurement conditions. On the PMSM model, SOO demonstrates accurate and repeatable parameter estimation within the adopted simulation framework.
Global-best-guided electric eel foraging optimizer for robust parameter identification of Lorenz and memristive chaotic systems
(Nature Research, 2026) İzci, Davut; Ekinci, Serdar; Ökten, İrfan; Tümen, Vedat; Bektaş Güneş, Burcu; Rashdan, Mostafa; Salman, Mohammad
Accurate parameter identification in chaotic dynamical systems constitutes a challenging inverse problem due to extreme sensitivity to initial conditions, pronounced nonlinearity, and highly multimodal error landscapes. To address these challenges, this study proposes a global-best-guided electric eel foraging optimization algorithm (g-EEFO), which enhances the original EEFO framework by embedding a behavior-aware and phase-dependent global learning mechanism. Unlike existing EEFO variants that rely solely on stochastic foraging dynamics, g-EEFO integrates global-best information as a soft cooperative signal that modulates the interacting, resting, hunting, and migrating behaviors without overriding them. In this way, global guidance acts as a directional bias rather than a dominant attractor, preserving ecological diversity while strengthening convergence coherence. For the first time, EEFO and its improved variant are applied to chaotic system parameter estimation. The proposed method is evaluated on two representative models: the classical Lorenz system and a structurally richer memristive chaotic system. Extensive numerical experiments, including statistical analysis, convergence profiling, boxplot distributions, and parameter-evolution trajectories, demonstrate the clear superiority of g-EEFO over several state-of-the-art metaheuristics. For the Lorenz system, g-EEFO achieves a best mean squared error of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:7.02\times\:{10}<^>{-26}$$\end{document}, which is six to twenty orders of magnitude lower than competing methods, while maintaining an exceptionally small standard deviation (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:4.58\times\:{10}<^>{-20}$$\end{document}). For the memristive system, g-EEFO attains a best error of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:8.19\times\:{10}<^>{-19}$$\end{document}, again outperforming all benchmarks by several orders of magnitude and exhibiting the highest run-to-run stability. In both cases, the estimated parameters match the true system values with near-perfect precision. These results confirm that the proposed behavior-aware global guidance fundamentally reshapes the search dynamics of EEFO, yielding substantial gains in convergence stability, numerical accuracy, and robustness. The g-EEFO therefore provides a powerful and reliable alternative for chaotic parameter identification and nonlinear system reconstruction across diverse dynamical regimes.
Double materiality and SWOT analysis: a practical framework for environmental impact and sustainability assessment
(Inderscience Enterprises Ltd, 2026) Yurtsever, Özlem; Çelik, Mustafa Cem; Uyar, Tanay Sıdkı
Double materiality is becoming a necessity in sustainability reporting, as it is essential for aligning organisational strategies with ESG priorities. This study proposes a framework to embed SWOT analysis within sustainability reporting, focusing on double materiality. By positioning SWOT as a bridge, the framework enables a holistic evaluation of internal capabilities and external ESG-related risks and opportunities. It encourages periodic updates to question sets, supporting an evolving, stakeholder-inclusive process aligned with annual reporting cycles. This approach offers organisations a practical reference to develop transparent, credible, and adaptable sustainability disclosures, reinforcing their strategic decision-making and compliance with increasingly complex reporting standards.
Hydrothermally synthesized chitosan-SnO2-CuO nanohybrids for antimicrobial surface engineering of GFRC
(Springer, 2026) Ramazanoglu, Doğu; Kamuran Duran, Pelin; Şahin, İdris; Subaşı, Serkan; Maraşlı, Muhammed
Glass fiber reinforced concrete (GFRC) panels are widely used in architectural and structural applications due to their low weight and design flexibility, but their hydrophilic and porous surfaces are prone to microbial colonization, compromising durability and aesthetics. In this study, GFRC surfaces were functionalized with hydrothermally synthesized chitosan-SnO2-CuO nanohybrids to enhance their physicochemical and antimicrobial properties. Structural and morphological analyses (XRD, SEM, FTIR, BET) confirmed successful nanohybrid incorporation, revealing increased surface area (3.40 to 15.77 m & sup2;/g) and modified chemical bonding. Water contact angle measurements indicated improved hydrophobicity (30.82 degrees to 63.49 degrees), while TGA/DTA showed enhanced thermal stability (final residue 16.9% to 18.07% at 800 degrees C). The nanohybrid coatings exhibited significant antimicrobial activity against E. coli, S. aureus, C. albicans, and T. tonsurans, with inhibition zones up to 36 mm and 29 mm at 2-3% additive concentrations. Thermal conductivity increased from 2.44 to 2.82 W/mK, demonstrating multifunctionality. These results highlight the potential of Ch-SnO2-CuO nanohybrid coatings as a robust, multifunctional strategy for producing antimicrobial, thermally stable GFRC surfaces suitable for hygiene-critical environments.
A data-driven h-infinity controller design with non-common lyapunov matrices for the active structural control having saturated actuators
(Institute of Electrical and Electronics Engineers Inc., 2025) Görmüş, Bilal; Yazıcı, Hakan; Küçükdemiral, İbrahim Beklan
This paper presents a data-driven H-infinity controller for active vibration control in structural systems having saturated actuators. The data-driven approach addresses parameter uncertainties by eliminating the need for system identification. The full-block S-procedure is used to formulate a convex optimization problem in the form of linear matrix inequalities (LMIs), though additional constraints may introduce conservatism. To mitigate this, the dilation technique with non-common Lyapunov matrices is employed, reducing conservatism and achieving a 12.7% lower H-infinity norm compared to common Lyapunov matrices. A seismically excited three-storey structure is used to validate the method. Simulations based on real-time data from the Kobe earthquake show that the proposed synthesis effectively reduces vibrations while control inputs never become saturated.
Calorimetric energetics and localized strain recovery in electron beam powder-bed fusion built NiTi shape memory alloy
(Elsevier, 2026) Sabirov, Tymur; Dilibal, Savaş; Lanba, Asheesh; Maris, Stella; Sfirri, Drew; Peduk, Gözde
Nickel-Titanium (NiTi or NiTiNOL) is widely recognized for its shape memory effect recovery, which is governed by the thermoelastic martensitic phase transformation. Its limited machinability has restricted its widespread adoption in applications requiring complex geometries. Additive manufacturing is attractive, but ubiquitous laser-based methods result in porous and micro-cracked microstructures that require extensive post-built thermal treatments to barely show functionality. Electron beam powder bed fusion (EB-PBF) presents a viable alternative, enabling near-net-shape fabrication while preserving the material's functional integrity due to high build temperatures and vacuum. This work quantifies the thermoelastic martensitic transformation energetics and links them to local deformation and recovery behavior in as-printed and annealed EB-PBF NiTi. Martensitic transformation behavior is analyzed using differential scanning calorimetry (DSC), which characterizes the transformation temperatures and enthalpies, which are then used to determine the elastic strain energy and irreversible/frictional energy for multiple cycles. We then use digital image correlation (DIC) to map localized reversible deformations that are responsible for the shape memory recovery. The results indicate that high-temperature annealing worsens thermoelastic potential, with as-printed samples exhibiting higher elastic strain energy. We also report large reversible localized deformation islands of detwinned martensite in a slurry of elastically deformed twinned martensite that are ultimately responsible for the global deformation that is then recovered upon heating, returning to its original shape for the as-printed material. The annealed sample on the other hand evolves a localized region of large plastic deformation that results in the fracture of the sample prior to useful detwinning of martensite taking place. These findings reinforce the role of DSC in quantifying phase transformation energetics and underscore EB-PBF's potential in creating NiTi components, without the need for post-annealing, for shape memory-based active applications.
Privacy-preserving VPN handshakes with Schnorr-based zero-knowledge proofs
(Elsevier Ltd, 2026) Yüce, Mehmet Fatih; Ertürk, Mehmet Ali; Aydın, Muhammed Ali
Modern Virtual Private Network (VPN) protocols rely on public-key-based handshakes that authenticate peers but can inadvertently reveal identifying or linkable information across sessions or network observers. This paper presents a privacy-preserving handshake framework that integrates Schnorr-based zero-knowledge proofs into existing VPN key-exchange mechanisms, allowing each party to prove key ownership without disclosing longterm identifiers such as static public keys. The framework is expressed as a generic extension layer applicable to a wide class of VPN protocols employing Diffie-Hellman-based mutual authentication (e.g., IKEv2/IPsec, OpenVPN, and WireGuard). To demonstrate feasibility, we integrate the scheme into WireGuard as a case study, yielding WireGuard-ZK. Implementation results show that the added privacy protection incurs modest computational and latency overhead while maintaining WireGuard's lightweight performance characteristics. The proposed design thus provides a generalizable cryptographic handshake model for privacy-preserving VPNs, combining theoretical soundness with practical deployability across modern tunneling frameworks.
Diabetic retinopathy classification on fundus images using YOLOv11 and CNN-based architectures
(Institute of Electrical and Electronics Engineers Inc., 2026) Akça, Alican; Köprülü, Elif Nur; Bektaş Güneş, Burcu; Kaçar, Fırat
Diabetic Retinopathy (DR) is a leading cause of vision impairment globally, driven by the rising prevalence of diabetes. Early and accurate diagnosis of DR through fundus imaging is crucial to prevent irreversible vision loss. Traditional manual examination methods are time-consuming, costly, and prone to inter-observer variability, creating a pressing need for automated solutions. In this study, we propose a deep learning-based approach for the classification of DR severity levels using fundus images. We evaluated several state-of-theart convolutional neural network (CNN) architectures-including VGG-16, MobileNetV2, EfficientNetB0, ResNet-50, VGG-19, and ConvNeXt-XLarge-and YOLOv11 for classification tasks. these models were trained and tested on annotated fundus datasets to classify DR into different severity stages, namely Grade 0 (No DR), Grade 1 (Mild DR), Grade 2 (Moderate DR), Grade 3 (Severe Non-Proliferative DR), and Grade 4 (Proliferative DR). Among these, YOLOv11 achieved the highest classification performance, demonstrating superior ability in severity grading.
Retrieval-augmented generation over low-structure media: news from tv and digitized newspapers
(Institute of Electrical and Electronics Engineers Inc., 2025) Uçar, Bilal Emir; Buluz Kömeçoǧlu, Başak; Güven, Ramazan; Coşkun, Ali Kemal; Kömeçoǧlu, Yavuz
This paper presents a practical and end-to-end implementation of a multimodal Retrieval-Augmented Generation (RAG) pipeline that integrates two traditionally underutilized but information-rich modalities in the news domain: transcribed television broadcasts and digitized digitized newspapers. While RAG has shown significant success in domains with clean digital text, its application to traditional media remains limited due to the inherent challenges in processing unstructured, noisy, and layout-complex content. To address these challenges, we propose a layout-aware document ingestion pipeline for digitized newspapers, powered by a semantic segmentation model trained. For television broadcasts, we integrate an optimized automatic speech recognition (ASR) and chunking framework. Both modalities are indexed under a unified hybrid retrieval architecture, combining dense and sparse representations to support accurate and semantically rich document retrieval. The system is deployed entirely on-premise and evaluated on a proprietary Turkish news corpus using standard retrieval metrics across both modalities.

Güncel Gönderiler