Engineering reports : open access最新文献

{"title":"Optimization of Tensile Strength Behavior in 3D-Printed PLA With 10% Terminalia chebula Nanocomposites: Influence of Strain Rate, Orientation, and Infill Percentage","authors":"R. Saravanan,&nbsp;Muthuselvan Balasubramanian,&nbsp;T. Sathish,&nbsp;Jayant Giri,&nbsp;A. Johnson Santhosh,&nbsp;Taoufik Saidani,&nbsp;Bashar Tarawneh","doi":"10.1002/eng2.70148","DOIUrl":"https://doi.org/10.1002/eng2.70148","url":null,"abstract":"<p>This study aims to investigate and maximize the tensile strength behavior of polylactic acid (PLA) 90% and <i>Terminalia chebula</i> nanoparticle (TCNP) 10% composites fabricated using fused deposition modeling (FDM) under varying strain rates (3, 6, 9 mm/min), orientations (0°, 45°, 90°), and infill percentages (30%, 60%, 90%). The tensile strength was analyzed to assess the combined influence of these parameters on the mechanical performance of the composites. At a low strain rate of 3 mm/min, the composites exhibited the highest tensile strength due to enhanced molecular alignment and stress redistribution, achieving maximum values in the 0° orientation across all infill percentages. Increasing the strain rate reduced tensile strength, with the material transitioning from ductile to brittle failure, especially at 9 mm/min, where rapid deformation hindered molecular realignment. The 0° orientation consistently demonstrated superior tensile strength due to efficient load transfer along printed layers, while the 90° orientation exhibited the weakest performance, attributed to stress concentrations at interlayer bonds. Higher infill percentages, 60% and 90%, improved material density, enhancing tensile strength but diminishing under higher strain rates. The study highlights the optimal mechanical performance of a maximum tensile strength of 45.67 ± 2.28 MPa, which was achieved at 90% infill, 3 mm/min strain rate, with 0° orientation, making it suitable for load-bearing applications. The findings provide insights into the tailoring of 3D-printed PLA-TCNP composites for specific applications, balancing strength, ductility, and controlled failure mechanisms.</p>","PeriodicalId":72922,"journal":{"name":"Engineering reports : open access","volume":"7 5","pages":""},"PeriodicalIF":1.8,"publicationDate":"2025-05-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/eng2.70148","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144074695","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Optimization of Compressive Strength Properties in Fused Deposition Modeling 3D Printed PLA/HA Composites for Bone Tissue Engineering Applications","authors":"Shashwath Patil,&nbsp;T. Sathish,&nbsp;Nashwan Adnan Othman,&nbsp;Bashar Tarawneh,&nbsp;Taoufik Saidani","doi":"10.1002/eng2.70133","DOIUrl":"https://doi.org/10.1002/eng2.70133","url":null,"abstract":"<p>This study investigates the optimization of 3D-printed polylactic acid (PLA) and hydroxyapatite (HA) composites for biomedical applications, focusing on enhancing mechanical properties through process parameter optimization and surface modification. The response surface methodology (RSM), along with post hoc statistical validation using Tukey's HSD test, was employed to evaluate the influence of nozzle temperature (200°C–240°C), layer height (0.1–0.3 mm), and HA filler ratio (3–9 wt%) on the compressive strength of both untreated and chemically treated composites. Silane treatment was applied to HA to improve interfacial bonding, resulting in a 5%–7% increase in compressive strength compared to untreated samples. The optimal conditions (240°C, 9% HA, 0.3 mm layer thickness) yielded a maximum compressive strength of 75.35 MPa in treated composites and 71.42 MPa for untreated samples. Statistical analysis confirmed that layer thickness and HA content significantly influenced mechanical performance. Contour plots and 3D response surfaces were also incorporated to visualize parameter interactions. Comparison with other optimization techniques demonstrated that RSM effectively minimized experimental runs while achieving superior mechanical properties. These findings suggest that chemically modified PLA/HA composites are promising candidates for load-bearing biomedical applications.</p>","PeriodicalId":72922,"journal":{"name":"Engineering reports : open access","volume":"7 5","pages":""},"PeriodicalIF":1.8,"publicationDate":"2025-05-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/eng2.70133","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144074697","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Different Estimation Methods for the Unit Xgamma Distribution Using Ranked Set Sampling","authors":"Diaa S. Metwally,&nbsp;Amal S. Hassan,&nbsp;Ehab M. Almetwally,&nbsp;Laxmi Prasad Sapkota,&nbsp;Ahmed M. Gemeay,&nbsp;Mohammed Elgarhy","doi":"10.1002/eng2.70157","DOIUrl":"https://doi.org/10.1002/eng2.70157","url":null,"abstract":"<p>Ranked set sampling (RSS) is an efficient sampling method when ranking observations is easier than precise measurement. Unlike simple random sampling (SRS), RSS can reduce costs. The unit Xgamma distribution (UXGD), defined over the interval (0,1), effectively captures the characteristics of negatively skewed datasets. This study aims to comprehensively compare several estimation methods, including maximum likelihood, Anderson-Darling, Kolmogorov, ordinary least squares, Anderson-Darling left tail second order, Cramer-von-Mises, left tail Anderson-Darling, weighted least squares, maximum product spacing, right tail Anderson-Darling, and five types of minimum spacing distance for the UXGD parameter under both RSS and SRS techniques. Through extensive simulations, we evaluate the performance of these estimators using multiple criteria under both designs. We rank the estimators based on their performance under both sampling schemes. Simulation findings indicate that the maximum product spacing and maximum likelihood estimation methods are superior to alternative approaches for assessing the estimated quality of RSS and SRS, respectively. It is interesting to note that for both SRS and RSS datasets, the estimates revealed by our model satisfy the consistency property. With an increase in the sample size, the estimates approach the true parameter values. Furthermore, the results highlight the efficiency gains of RSS over SRS, as evidenced by improved accuracy metrics. Two real-world applications, including COVID-19 data from the United Kingdom and France, demonstrate the practical utility of our findings.</p>","PeriodicalId":72922,"journal":{"name":"Engineering reports : open access","volume":"7 5","pages":""},"PeriodicalIF":1.8,"publicationDate":"2025-05-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/eng2.70157","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144074628","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Numerical Simulation Study of CO2 Injection for Enhanced Recovery in Depleted Gas Reservoirs","authors":"Mengyu Wang,&nbsp;Yu Luo,&nbsp;Zihan Zhao,&nbsp;Yulong Zhao,&nbsp;Yu Li,&nbsp;Xia Wang,&nbsp;Nan Qin,&nbsp;Cheng Cao,&nbsp;Ruoyu Mu","doi":"10.1002/eng2.70165","DOIUrl":"https://doi.org/10.1002/eng2.70165","url":null,"abstract":"<p>The Sichuan Basin is rich in gas reservoirs, most of which are in the middle to late stages of development and have significant potential for CCUS (Carbon Capture, Utilization, and Storage). To reduce CO₂ emissions and assess the feasibility of CCUS-EGR, Wells 47 and 67 in the Maokou Formation of the Wolonghe Gas Field were selected for CO₂ injection tests. Drawing on gas storage operation and reservoir development experiences, numerical simulation was used to design parameters such as CO₂ injection mode, timing, and injection-production well pattern. Simulations were conducted on the CO₂ storage mechanism in the depleted gas reservoir of Well 47 and the numerical model study of CO₂ injection and storage at Well 67. Firstly, Well 47 was analyzed to assess changes in reservoir temperature and pressure under varying CO₂ injection rates, timings, and temperatures, revealing the CO₂ migration distribution within the formation. Then, a comparative analysis of Well 67 evaluated cumulative natural gas production under different injection gases, production rates, and injection-production ratios, involving Wells 67, 83, and 47. Finally, the optimal injection-production parameters and CO₂ injection scheme were determined: (1) Well 067-C1 and Well 067-C2 were injected with CO₂ at 150,000 m³/day; (2) Well 47, Well 67, and Well 83 produce gas at 100,000 m³/day; (3) Well 83 and Well 47 were closed immediately after the breakthrough. The scheme can cumulatively recover 530 million m³ of natural gas, accounting for 9.64% of dynamic reserves, accounting for 8.03% of geological reserves, and can store 3.686 million tons of CO₂. This simulation test is the first new energy project of Southwest Oil &amp; Gas Field Company on CO₂ injection to enhance oil recovery. This simulation test provides support for the gas field company to explore the scientific development path of low-carbon new energy through the pilot test of CCUS-EGR.</p>","PeriodicalId":72922,"journal":{"name":"Engineering reports : open access","volume":"7 5","pages":""},"PeriodicalIF":1.8,"publicationDate":"2025-05-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/eng2.70165","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144074696","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"STSA-Based Early-Stage Detection of Small Brain Tumors Using Neural Network","authors":"Nafiul Hasan,&nbsp;Md. Masud Rana,&nbsp;Md Mahmudul Hasan,&nbsp;AKM Azad,&nbsp;Dil Afroz,&nbsp;Md Mostafizur Rahman Komol,&nbsp;Mousumi Aktar,&nbsp;Mohammad Ali Moni","doi":"10.1002/eng2.70135","DOIUrl":"https://doi.org/10.1002/eng2.70135","url":null,"abstract":"<p>Early-stage brain tumor detection is critical for improving patient outcomes, optimizing treatment strategies, and enhancing healthcare resource allocation. However, existing state-of-the-art techniques struggle to detect tumors smaller than 5 mm due to their minimal dimensions and complex electromagnetic interactions. This study introduces a machine learning-based classification approach for early-stage Astrocytoma tumors (grades I and II) using step-constant tapered slot antenna (STSA) parameters. By leveraging scattering (S), admittance (Y), and impedance (Z) parameters as input features, an Artificial Neural Network (ANN) achieved a 99.95% classification accuracy for tumors with radii of 3 mm and 5 mm. Among the input features, impedance (Z) was identified as the most significant contributor to classification accuracy, whereas the S-parameter exhibited the lowest performance at 84.21% accuracy. The proposed methodology was benchmarked against Support Vector Machine (SVM), K-Nearest Neighbor (KNN), Random Forest Classifier (RFC), and Graph Convolutional Neural Network (GCN), demonstrating superior classification performance across different tumor sizes. Additionally, the system maintained a low Specific Absorption Rate (SAR) of 0.30 W/Kg, reinforcing its suitability for biomedical antenna-based applications. An ablation study further confirmed that Z₂₂ and Z₁₄ phase components within the impedance matrix were particularly influential, as revealed through Local Interpretable Model-Agnostic Explanations (LIME), an explainable AI (XAI) technique. The proposed method was evaluated using a publicly available dataset, validating its robustness. These findings highlight the potential of STSA-based machine learning models for accurate, non-invasive early-stage brain tumor classification, enabling cost-effective, scalable diagnostics.</p>","PeriodicalId":72922,"journal":{"name":"Engineering reports : open access","volume":"7 5","pages":""},"PeriodicalIF":1.8,"publicationDate":"2025-05-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/eng2.70135","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144074257","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Correction to “Metallographic Spheroidization Rate Classification by Using Deep Learning”","authors":"","doi":"10.1002/eng2.70182","DOIUrl":"https://doi.org/10.1002/eng2.70182","url":null,"abstract":"<p>L. Chiu-Chin, P.-H. Chiang, K.-H. Chen, Y.-J. Pan, C.-H. Chen, W. Yi-Shun, J.-C. Lee, “Metallographic Spheroidization Rate Classification by Using Deep Learning,” <i>Engineering Reports</i>, 7 (2025): e70081, https://doi.org/10.1002/eng2.70081.</p><p>In the originally published article, the current list of Author and their affiliations were as follows:</p><p>Lin Chiu-Chin¹ | Pei-Hsuan Chiang² | Ke-Hao Chen² | Yu-Jen Pan³ | Chung-Hsien Chen² | Wang Yi-Shun⁴ | Jen-Chun Lee¹</p><p>¹Department of Telecommunication Engineering, National Kaohsiung University of Science and Technology, Kaohsiung, Taiwan</p><p>²Metal Industries Research and Development Centre (MIRDC), Kaohsiung, Taiwan</p><p>³Department of Supply Chain Management, National Kaohsiung University of Science and Technology, Kaohsiung, Taiwan</p><p>⁴Ph.D. Program in Maritime Science and Technology, National Kaohsiung University of Science and Technology, Kaohsiung, Taiwan</p><p>Upon checking, it was discovered that the name order of the 1st and 6th authors was incorrect. The order should be as follows:</p><p>Chiu-Chin Lin</p><p>Yi-Shun Wang</p><p>Furthermore, it was discovered that Pei-Hsuan Chiang is affiliated with the below:</p><p>⁴Ph.D. Program in Maritime Science and Technology, National Kaohsiung University of Science and Technology, Kaohsiung, Taiwan</p><p>Therefore, the name order of the Authors and affiliations were now revised to:</p><p>Chiu-Chin Liu¹ | Pei-Hsuan Chiang⁴ | Ke-Hao Chen² | Yu-Jen Pan³ | Chung-Hsien Chen² | Yi-Shun Wang⁴ | Jen-Chun Lee¹</p><p>The affiliation list remains the same.</p><p>We apologize for this error.</p>","PeriodicalId":72922,"journal":{"name":"Engineering reports : open access","volume":"7 5","pages":""},"PeriodicalIF":1.8,"publicationDate":"2025-05-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/eng2.70182","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143949748","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Correction to “Adaptive Ensemble Framework With Synthetic Sampling for Tackling Class Imbalance Problem”","authors":"","doi":"10.1002/eng2.70181","DOIUrl":"https://doi.org/10.1002/eng2.70181","url":null,"abstract":"<p>R. Sasirekha and B. Kanisha, “Adaptive Ensemble Framework With Synthetic Sampling for Tackling Class Imbalance Problem,” <i>Engineering Reports</i> 7 (2025): e70109, https://doi.org/10.1002/eng2.70109.</p><p>In the originally published article, the Authors were affiliated to Department of Computing Technologies, Faculty of Engineering and Technology, SRM Institute of Science and Technology, Chengulpattu, India.</p><p>However, it was discovered that the address is incomplete.</p><p>The corrected affiliation of the Authors should be:</p><p>Department of Computing Technologies, Faculty of Engineering and Technology, SRM Institute of Science and Technology, Kattankulathur Campus, Chengulpattu, India.</p><p>We apologize for this error.</p>","PeriodicalId":72922,"journal":{"name":"Engineering reports : open access","volume":"7 5","pages":""},"PeriodicalIF":1.8,"publicationDate":"2025-05-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/eng2.70181","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143949747","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Review of Lithium-Ion Battery Empirical and Semi-Empirical Aging Models for Off-Grid Renewable Energy Systems Application","authors":"Isaac Gwayi,&nbsp;Sarah Paul Ayeng'o,&nbsp;Cuthbert Z. M. Kimambo","doi":"10.1002/eng2.70169","DOIUrl":"https://doi.org/10.1002/eng2.70169","url":null,"abstract":"<p>Aging of lithium-ion (Li-ion) batteries in off-grid renewable energy systems (RESs) can be monitored and controlled using battery management systems (BMSs) which utilize battery aging models. Empirical and semi-empirical models (EMs) of battery aging are preferred for BMSs due to their simplicity and intuitiveness. This study is unique as it aims at identifying appropriate empirical and semi-EMs, in terms of complexity and current fluctuation representation, for BMS for off-grid RESs. Different EMs of Li-ion battery calendar and cycle aging have been extracted from literature and compared mainly in terms of complexity, current fluctuation representation, and modeling of capacity fade and resistance increase. The extracted models have been put in groups which are based on modeling format used, namely: calendar aging (only) models (CAOM), cycle aging (only) models (CYAOM), calendar and cycle aging (separated) models (CCYAOM), and calendar and cycle aging (combined) models (CCYACM). Results show that three models meet requirements for BMS for off-grid RESs. The three models fall under CYAOM, CCYAOM, and CCYACM. The study further finds that 54% of EMs model current fluctuation as an aging factor, 92% model aging in terms of capacity fade, and 46% model aging as resistance increase. Furthermore, the study recommends comparison of EMs through simulations to further validate the different listed models. It also recommends evaluation of the models to establish an appropriate way of representing Li-ion battery aging, whether in terms of capacity fade or resistance increase.</p>","PeriodicalId":72922,"journal":{"name":"Engineering reports : open access","volume":"7 5","pages":""},"PeriodicalIF":1.8,"publicationDate":"2025-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/eng2.70169","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143925794","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Balancing Efficiency and Emissions: A Study on Biodiesel-Hydrogen Dual-Fuel Engine Performance Under Varying Injection Timings","authors":"Vasanthkumar Periyathambi,&nbsp;Manikandan Ezhumalai,&nbsp;Mohan Govindasamy,&nbsp;Ratchagaraja Dhairiyasamy,&nbsp;Deekshant Varshney,&nbsp;Subhav Singh","doi":"10.1002/eng2.70174","DOIUrl":"https://doi.org/10.1002/eng2.70174","url":null,"abstract":"<p>The growing demand for cleaner energy sources has intensified research on alternative fuels for diesel engines to mitigate environmental concerns. This study presents a novel approach by investigating the combined effect of injection timing (IT) and hydrogen enrichment on the performance, combustion, and emissions of a single-cylinder, four-stroke variable compression ratio (VCR) engine fueled with a CI20 biodiesel-hydrogen dual-fuel blend. Injection timings of IT24, IT27, IT30, and IT33 degrees before TDC and hydrogen flow rates of 4, 8, 12, and 16 LPM were tested. The research showed that IT30 operating with 16 LPM hydrogen flow produced the maximum brake thermal efficiency (BTE) of 32.7% and the minimum brake-specific fuel consumption (BSFC) of 0.26 kJ/kWh while working at full load. The experimental conditions led to 7.7% higher cylinder pressure and an 8% enhancement of the heat release rate when compared with basic operation. The emission analysis showed that carbon monoxide (CO) decreased by 48%, hydrocarbons (HC) decreased by 22.2%, and the smoke opacity diminished by 34.3%, while NO_<i>x</i> emissions rose by 6.3% to 755 ppm due to higher combustion temperatures. Response Surface Methodology (RSM) optimization found IT30.25° should be used with 16 LPM H₂ as the best operating condition, which forecasts a brake thermal efficiency of 32.52% and NO_<i>x</i> output at 751 ppm. This work brings an innovative approach by combining biodiesel combustion enhancement through precise control of hydrogen flow and timing adjustments, which received statistical validation through diagnostic methods. The obtained data create a performance-enhanced pathway to improve efficiency under BS-VI and Euro VI emission standards for CI20-hydrogen blends, which demonstrate potential as efficient decarbonized engine technology solutions.</p>","PeriodicalId":72922,"journal":{"name":"Engineering reports : open access","volume":"7 5","pages":""},"PeriodicalIF":1.8,"publicationDate":"2025-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/eng2.70174","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143925820","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Optimizing Deep Learning Models for Resource-Constrained Environments With Cluster-Quantized Knowledge Distillation","authors":"Niaz Ashraf Khan,&nbsp;A. M. Saadman Rafat","doi":"10.1002/eng2.70187","DOIUrl":"https://doi.org/10.1002/eng2.70187","url":null,"abstract":"<p>Deep convolutional neural networks (CNNs) are highly effective in computer vision tasks but remain challenging to deploy in resource-constrained environments due to their high computational and memory requirements. Conventional model compression techniques, such as pruning and post-training quantization, often compromise model accuracy by decoupling compression from training. Furthermore, traditional knowledge distillation approaches rely on full-precision teacher models, limiting their effectiveness in compressed settings. To address these issues, we propose Cluster-Quantized Knowledge Distillation (CQKD), a novel framework that integrates structured pruning with knowledge distillation, incorporating cluster-based weight quantization directly into the training loop. Unlike existing methods, CQKD applies quantization to both the teacher and student models, ensuring a more effective transfer of compressed knowledge. By leveraging layer-wise K-means clustering, our approach achieves extreme model compression while maintaining high accuracy. Experimental results on CIFAR-10 and CIFAR-100 demonstrate the effectiveness of CQKD, achieving compression ratios of 34,000× while preserving competitive accuracy—97.9% on CIFAR-10 and 91.2% on CIFAR-100. These results highlight the feasibility of CQKD for efficient deep learning model deployment in low-resource environments.</p>","PeriodicalId":72922,"journal":{"name":"Engineering reports : open access","volume":"7 5","pages":""},"PeriodicalIF":1.8,"publicationDate":"2025-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/eng2.70187","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143925990","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}