Pub Date : 2025-11-28DOI: 10.1016/j.asej.2025.103886
Ju Li , Junhong Chen , Fengjun Shang
In this paper, we investigate the problem of dynamic event-triggered consensus tracking control for nonlinear multi-agent systems (MASs) with actuator failures and unknown dead zones. The MASs involve mismatched unknown parameters, actuator failures, and unknown dead zones, which make the consensus control problem difficult. To this end, by using the bounded estimation method, smoothing function method, and adaptive technique, a distributed adaptive fault-tolerant control (FTC) scheme based on the backstepping technique is designed, which compensates adaptively for effects of actuator failures and unknown dead zones. Moreover, a new dynamic event-triggered control (DETC) strategy is developed to reduce the communication burden in contrast to the existing static event-triggered control. Based on the Lyapunov method, it is shown that the consensus tracking control can be achieved and the Zeno behavior does not occur. Finally, the effectiveness of the proposed control method is validated through two simulation examples.
{"title":"Dynamic event-triggered consensus tracking control for nonlinear multi-agent systems with actuator failures and unknown dead zones","authors":"Ju Li , Junhong Chen , Fengjun Shang","doi":"10.1016/j.asej.2025.103886","DOIUrl":"10.1016/j.asej.2025.103886","url":null,"abstract":"<div><div>In this paper, we investigate the problem of dynamic event-triggered consensus tracking control for nonlinear multi-agent systems (MASs) with actuator failures and unknown dead zones. The MASs involve mismatched unknown parameters, actuator failures, and unknown dead zones, which make the consensus control problem difficult. To this end, by using the bounded estimation method, smoothing function method, and adaptive technique, a distributed adaptive fault-tolerant control (FTC) scheme based on the backstepping technique is designed, which compensates adaptively for effects of actuator failures and unknown dead zones. Moreover, a new dynamic event-triggered control (DETC) strategy is developed to reduce the communication burden in contrast to the existing static event-triggered control. Based on the Lyapunov method, it is shown that the consensus tracking control can be achieved and the Zeno behavior does not occur. Finally, the effectiveness of the proposed control method is validated through two simulation examples.</div></div>","PeriodicalId":48648,"journal":{"name":"Ain Shams Engineering Journal","volume":"17 1","pages":"Article 103886"},"PeriodicalIF":5.9,"publicationDate":"2025-11-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145614487","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-27DOI: 10.1016/j.asej.2025.103877
Sana Yasin , Umar Draz , Hazem M. El-Hageen , Tariq Ali , Yousef H. Alfaifi , Muhammad Ayaz , Low Tang Jung , El-Hadi M. Aggoune
The rapid advancement of Industry 4.0 has driven the need for intelligent, responsive manufacturing systems that can anticipate and mitigate disruptions, reduce downtime, and optimize performance. However, traditional manufacturing systems often rely on reactive maintenance strategies, leading to unexpected downtimes and increased operational costs. These limitations necessitate a shift toward predictive optimization for real-time condition monitoring and proactive failure prevention. This study presents a Tensor-Based Semantic Digital Twin (SDT) framework that integrates a hybrid CNN-LSTM model with ontology-based decision-making to provide a context-aware predictive optimization system. The SDT framework outperforms traditional methods, such as Deep Learning-Based Predictive Maintenance (DL-PdM) and Hybrid Anomaly Detection in Manufacturing (HAD-M), by significantly reducing equipment downtime and lowering costs per unit while maintaining high predictive accuracy and operational efficiency. The study demonstrates that SDTs provide real-time operational intelligence, ensuring manufacturing systems are more resilient, sustainable, and cost-effective. Unlike traditional approaches that rely on static models, the proposed framework enables adaptive learning from evolving sensor data, allowing manufacturing facilities to anticipate failures proactively and optimize resource utilization dynamically. By leveraging a hybrid CNN-LSTM model, the framework accurately predicts equipment failures, ensuring a 30% reduction in unexpected downtimes and a 20% improvement in energy consumption efficiency. Additionally, ontology-based decision-making enables context-aware automation, leading to an 18% decrease in maintenance costs and adaptive load balancing in dynamic industrial environments. Additionally, the proposed system extends beyond predictive maintenance by incorporating automated reactive control, making real-time adjustments to minimize inefficiencies. The findings establish a new standard for intelligent manufacturing, ensuring enhanced resilience, adaptability, and cost efficiency of smart manufacturing.
{"title":"Enhancing smart manufacturing: a tensor-based ontology framework for predictive optimization using semantic digital twin","authors":"Sana Yasin , Umar Draz , Hazem M. El-Hageen , Tariq Ali , Yousef H. Alfaifi , Muhammad Ayaz , Low Tang Jung , El-Hadi M. Aggoune","doi":"10.1016/j.asej.2025.103877","DOIUrl":"10.1016/j.asej.2025.103877","url":null,"abstract":"<div><div>The rapid advancement of Industry 4.0 has driven the need for intelligent, responsive manufacturing systems that can anticipate and mitigate disruptions, reduce downtime, and optimize performance. However, traditional manufacturing systems often rely on reactive maintenance strategies, leading to unexpected downtimes and increased operational costs. These limitations necessitate a shift toward predictive optimization for real-time condition monitoring and proactive failure prevention. This study presents a Tensor-Based Semantic Digital Twin (SDT) framework that integrates a hybrid CNN-LSTM model with ontology-based decision-making to provide a context-aware predictive optimization system. The SDT framework outperforms traditional methods, such as Deep Learning-Based Predictive Maintenance (DL-PdM) and Hybrid Anomaly Detection in Manufacturing (HAD-M), by significantly reducing equipment downtime and lowering costs per unit while maintaining high predictive accuracy and operational efficiency. The study demonstrates that SDTs provide real-time operational intelligence, ensuring manufacturing systems are more resilient, sustainable, and cost-effective. Unlike traditional approaches that rely on static models, the proposed framework enables adaptive learning from evolving sensor data, allowing manufacturing facilities to anticipate failures proactively and optimize resource utilization dynamically. By leveraging a hybrid CNN-LSTM model, the framework accurately predicts equipment failures, ensuring a 30% reduction in unexpected downtimes and a 20% improvement in energy consumption efficiency. Additionally, ontology-based decision-making enables context-aware automation, leading to an 18% decrease in maintenance costs and adaptive load balancing in dynamic industrial environments. Additionally, the proposed system extends beyond predictive maintenance by incorporating automated reactive control, making real-time adjustments to minimize inefficiencies. The findings establish a new standard for intelligent manufacturing, ensuring enhanced resilience, adaptability, and cost efficiency of smart manufacturing.</div></div>","PeriodicalId":48648,"journal":{"name":"Ain Shams Engineering Journal","volume":"17 1","pages":"Article 103877"},"PeriodicalIF":5.9,"publicationDate":"2025-11-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145614485","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-27DOI: 10.1016/j.asej.2025.103892
Yuqing Xia, Yingkang Yao, Yongsheng Jia, Jinshan Sun, Nan Jiang
Blasting demolition is widely used for removing high-rise buildings due to its safety and efficiency, but induced vibrations and collapse impacts can damage nearby structures. To study the dynamic behavior of an adjacent subway tunnel subjected to blasting and collapse impact loads during high-rise building blasting demolition, a 24-story building demolition project is taken as a research case. Vibrometers and strain gauges are installed in the tunnel to record its dynamic response. Additionally, an elaborate 3D numerical model is established to simulate the collapse process and clarify three distinct impact load characteristics. It is indicated that blasting vibration from the explosion and the high-velocity collapse of the building’s upper parts have negligible effects on the tunnel. However, the impact load generated on the basement, resulting from the failure of the primary incision and the building’s subsequent subsidence, exerts a critical influence on the tunnel. A parameter analysis reveals the decisive effect of the Er/Es ratio. It indicates that segment stress, unlike PPV, is the more reliable criterion for evaluating segment safety under such impacts.
{"title":"Dynamic behavior of an adjacent tunnel under blasting and collapse impact loads: a case study of high-rise building blasting demolition","authors":"Yuqing Xia, Yingkang Yao, Yongsheng Jia, Jinshan Sun, Nan Jiang","doi":"10.1016/j.asej.2025.103892","DOIUrl":"10.1016/j.asej.2025.103892","url":null,"abstract":"<div><div>Blasting demolition is widely used for removing high-rise buildings due to its safety and efficiency, but induced vibrations and collapse impacts can damage nearby structures. To study the dynamic behavior of an adjacent subway tunnel subjected to blasting and collapse impact loads during high-rise building blasting demolition, a 24-story building demolition project is taken as a research case. Vibrometers and strain gauges are installed in the tunnel to record its dynamic response. Additionally, an elaborate 3D numerical model is established to simulate the collapse process and clarify three distinct impact load characteristics. It is indicated that blasting vibration from the explosion and the high-velocity collapse of the building’s upper parts have negligible effects on the tunnel. However, the impact load generated on the basement, resulting from the failure of the primary incision and the building’s subsequent subsidence, exerts a critical influence on the tunnel. A parameter analysis reveals the decisive effect of the <em>E<sub>r</sub>/E<sub>s</sub></em> ratio. It indicates that segment stress, unlike PPV, is the more reliable criterion for evaluating segment safety under such impacts.</div></div>","PeriodicalId":48648,"journal":{"name":"Ain Shams Engineering Journal","volume":"17 1","pages":"Article 103892"},"PeriodicalIF":5.9,"publicationDate":"2025-11-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145614486","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Breast cancer is one of the most common diseases that affects women around the world. Finding it early can help save lives. Doctors usually use tests like biopsy, ultrasound, CT scan, and mammography to detect breast cancer. In this study, we created a computer-based method that helps in detecting breast cancer more accurately. The process has four main steps — preprocessing, segmentation, feature selection, and classification. First, we clean the mammogram images using a filter to remove unwanted noise. Next, we separate the important part of the image using a special method called Thresholding-Based Level Set. Then, we choose only the most important details (features) from the image using a new hybrid method called Improved Grey Wolf Optimization with Seagull Optimization Algorithm. Finally, we use a machine learning model named CatBoost to identify if the tumor is benign (non-cancerous) or malignant (cancerous). When tested on a dataset, our method showed excellent results with 99.2 % accuracy. This shows that our model can help doctors detect breast cancer early and more correctly.
{"title":"An improved hybrid feature selection and classification framework for breast cancer detection using mammography images","authors":"Aniruddha Deka , Debashis Dev Misra , Munsifa Firdaus Khan Barbhuyan , Mudassir Khan , Mostaque Md. Morshedur Hassan , Mohammed Ashfaq Hussain","doi":"10.1016/j.asej.2025.103897","DOIUrl":"10.1016/j.asej.2025.103897","url":null,"abstract":"<div><div>Breast cancer is one of the most common diseases that affects women around the world. Finding it early can help save lives. Doctors usually use tests like biopsy, ultrasound, CT scan, and mammography to detect breast cancer. In this study, we created a computer-based method that helps in detecting breast cancer more accurately. The process has four main steps — preprocessing, segmentation, feature selection, and classification. First, we clean the mammogram images using a filter to remove unwanted noise. Next, we separate the important part of the image using a special method called Thresholding-Based Level Set. Then, we choose only the most important details (features) from the image using a new hybrid method called Improved Grey Wolf Optimization with Seagull Optimization Algorithm. Finally, we use a machine learning model named CatBoost to identify if the tumor is benign (non-cancerous) or malignant (cancerous). When tested on a dataset, our method showed excellent results with 99.2 % accuracy. This shows that our model can help doctors detect breast cancer early and more correctly.</div></div>","PeriodicalId":48648,"journal":{"name":"Ain Shams Engineering Journal","volume":"17 1","pages":"Article 103897"},"PeriodicalIF":5.9,"publicationDate":"2025-11-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145614488","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper presents an advanced fault diagnosis framework for renewable energy systems by leveraging a novel Adaptive Polynomial Kolmogorov Arnold Network (Adaptive PolyKAN). The proposed method is evaluated on two distinct applications: a wind energy conversion system and a grid-connected photovoltaic (PV) system, each characterized by complex, nonlinear fault patterns. A comprehensive comparison is conducted against a range of classical and neural classifiers, including Random Forest (RF), Support Vector Machine (SVM), and others. Experimental results demonstrate that Adaptive PolyKAN consistently achieves superior classification accuracy, reaching 99.96 % for wind data and 95.61 % for PV data, outperforming conventional methods across all performance metrics. To improve computational efficiency, an autoencoder-based dimensionality reduction strategy is incorporated, resulting in a reduction of execution time by over 88 % and memory usage by 40 %, while preserving high diagnostic accuracy, maintaining 99.96 % on the wind data and increasing to 96.47 % on the PV data. The results confirm the robustness, adaptability, and efficiency of the proposed framework, highlighting its potential for intelligent fault diagnosis in complex renewable energy systems.
{"title":"Adaptive PolyKAN-based autoencoder for fault detection and classification in wind and solar power systems","authors":"Khadija Attouri , Majdi Mansouri , Abdelmalek Kouadri","doi":"10.1016/j.asej.2025.103884","DOIUrl":"10.1016/j.asej.2025.103884","url":null,"abstract":"<div><div>This paper presents an advanced fault diagnosis framework for renewable energy systems by leveraging a novel Adaptive Polynomial Kolmogorov Arnold Network (Adaptive PolyKAN). The proposed method is evaluated on two distinct applications: a wind energy conversion system and a grid-connected photovoltaic (PV) system, each characterized by complex, nonlinear fault patterns. A comprehensive comparison is conducted against a range of classical and neural classifiers, including Random Forest (RF), Support Vector Machine (SVM), and others. Experimental results demonstrate that Adaptive PolyKAN consistently achieves superior classification accuracy, reaching 99.96 % for wind data and 95.61 % for PV data, outperforming conventional methods across all performance metrics. To improve computational efficiency, an autoencoder-based dimensionality reduction strategy is incorporated, resulting in a reduction of execution time by over 88 % and memory usage by 40 %, while preserving high diagnostic accuracy, maintaining 99.96 % on the wind data and increasing to 96.47 % on the PV data. The results confirm the robustness, adaptability, and efficiency of the proposed framework, highlighting its potential for intelligent fault diagnosis in complex renewable energy systems.</div></div>","PeriodicalId":48648,"journal":{"name":"Ain Shams Engineering Journal","volume":"17 1","pages":"Article 103884"},"PeriodicalIF":5.9,"publicationDate":"2025-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145614484","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-23DOI: 10.1016/j.asej.2025.103882
Mohamed A. Ahmed, Mona A. Bayoumi
This study proposes a novel deep feedforward neural network (DFNN)-based control for voltage sag prediction and dynamic voltage restorer (DVR) integration with low-voltage ride-through (LVRT) to improve voltage stability and energy efficiency in wind turbine systems. Proposed approach precisely predicts sag duration, adaptively regulates DVR activation according to LVRT profile to avoid redundant compensation. This predictive control reduces DVR operating time and energy consumption while maintaining voltage stability. Synthetic datasets representing various grid conditions, including distorted voltage, enable DFNN to achieve a prediction accuracy near 99 %, a time error margin of 0.005–0.02 s and a response time of 0.016 s. This precise prediction improves DVR energy efficiency by nearly 30 % per long-duration fault. Compared to a support vector regression (SVR) model, DFFN achieves 33.3 % faster response and lower error metrics. Simulation results validated in a MATLAB/Simulink demonstrate effectiveness of proposed approach in enhancing LVRT capability and overall grid efficiency.
{"title":"A novel deep learning-based control for voltage sag prediction and DVR–LVRT coordination in grid-connected wind turbine systems","authors":"Mohamed A. Ahmed, Mona A. Bayoumi","doi":"10.1016/j.asej.2025.103882","DOIUrl":"10.1016/j.asej.2025.103882","url":null,"abstract":"<div><div>This study proposes a novel deep feedforward neural network (DFNN)-based control for voltage sag prediction and dynamic voltage restorer (DVR) integration with low-voltage ride-through (LVRT) to improve voltage stability and energy efficiency in wind turbine systems. Proposed approach precisely predicts sag duration, adaptively regulates DVR activation according to LVRT profile to avoid redundant compensation. This predictive control reduces DVR operating time and energy consumption while maintaining voltage stability. Synthetic datasets representing various grid conditions, including distorted voltage, enable DFNN to achieve a prediction accuracy near 99 %, a time error margin of 0.005–0.02 s and a response time of 0.016 s. This precise prediction improves DVR energy efficiency by nearly 30 % per long-duration fault. Compared to a support vector regression (SVR) model, DFFN achieves 33.3 % faster response and lower error metrics. Simulation results validated in a MATLAB/Simulink demonstrate effectiveness of proposed approach in enhancing LVRT capability and overall grid efficiency.</div></div>","PeriodicalId":48648,"journal":{"name":"Ain Shams Engineering Journal","volume":"17 1","pages":"Article 103882"},"PeriodicalIF":5.9,"publicationDate":"2025-11-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145614483","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-21DOI: 10.1016/j.asej.2025.103871
Maedeh GholamAzad, Alireza Eydi
As global industrial operations expand, the complexity and volume of suppliers have increased, making sustainable supplier selection (SSS) a strategic imperative for resilient supply chains (SCs). Traditional evaluation methods often fail to handle large datasets and dynamic sustainability metrics, resulting in suboptimal decisions. This study introduces a novel hybrid intelligent framework that integrates Most Productive Scale Size Data Envelopment Analysis (MPSS-DEA) with three Artificial intelligence (AI) algorithms—Artificial Neural Networks (ANNs), K-Nearest Neighbors (KNN), and Chi-squared Automatic Interaction Detection (CHAID)—to enhance both accuracy and scalability in supplier evaluation. Applied to 362 suppliers and 38 sustainability criteria in the petrochemical industry, the proposed framework achieved high classification accuracies of 92.5% (DEA-CHAID), 91.9% (DEANN), and 91.4% (DEA-KNN). The model also demonstrated strong discrimination power, with ROC AUC scores of 0.96, 0.95, and 0.94, respectively. Predictor importance analysis revealed that some of the features, such as R2, SE2, QA4, CA6, and DE3, were the most influential features across all models. Beyond performance metrics, the framework offers real-time supplier replacement, reduced computational complexity, and modular adaptability across industries. It supports ethical and sustainable sourcing by integrating economic, environmental, and social dimensions into decision-making. The intelligent architecture enables lifecycle analysis, promotes transparency, and aligns with global sustainability standards. This research contributes a scalable, interpretable, and data-driven solution for sustainable supplier selection, bridging the gap between traditional DEA models and modern artificial intelligence (AI) techniques. Its applicability across diverse industrial contexts positions it as a robust tool for strategic procurement and supply chain resilience.
{"title":"A hybrid data envelopment analysis and artificial intelligence framework for sustainable supplier selection: a case study in the petrochemical industry","authors":"Maedeh GholamAzad, Alireza Eydi","doi":"10.1016/j.asej.2025.103871","DOIUrl":"10.1016/j.asej.2025.103871","url":null,"abstract":"<div><div>As global industrial operations expand, the complexity and volume of suppliers have increased, making sustainable supplier selection (SSS) a strategic imperative for resilient supply chains (SCs). Traditional evaluation methods often fail to handle large datasets and dynamic sustainability metrics, resulting in suboptimal decisions. This study introduces a novel hybrid intelligent framework that integrates Most Productive Scale Size Data Envelopment Analysis (MPSS-DEA) with three Artificial intelligence (AI) algorithms—Artificial Neural Networks (ANNs), K-Nearest Neighbors (KNN), and Chi-squared Automatic Interaction Detection (CHAID)—to enhance both accuracy and scalability in supplier evaluation. Applied to 362 suppliers and 38 sustainability criteria in the petrochemical industry, the proposed framework achieved high classification accuracies of 92.5% (DEA-CHAID), 91.9% (DEANN), and 91.4% (DEA-KNN). The model also demonstrated strong discrimination power, with ROC AUC scores of 0.96, 0.95, and 0.94, respectively. Predictor importance analysis revealed that some of the features, such as R2, SE2, QA4, CA6, and DE3, were the most influential features across all models. Beyond performance metrics, the framework offers real-time supplier replacement, reduced computational complexity, and modular adaptability across industries. It supports ethical and sustainable sourcing by integrating economic, environmental, and social dimensions into decision-making. The intelligent architecture enables lifecycle analysis, promotes transparency, and aligns with global sustainability standards. This research contributes a scalable, interpretable, and data-driven solution for sustainable supplier selection, bridging the gap between traditional DEA models and modern artificial intelligence (AI) techniques. Its applicability across diverse industrial contexts positions it as a robust tool for strategic procurement and supply chain resilience.</div></div>","PeriodicalId":48648,"journal":{"name":"Ain Shams Engineering Journal","volume":"17 1","pages":"Article 103871"},"PeriodicalIF":5.9,"publicationDate":"2025-11-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145568883","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-21DOI: 10.1016/j.asej.2025.103868
Aleeza Kiran , Muhammad Yaseen , Aziz Khan , Thabet Abdeljawad , Manar A. Alqudah , Rajermani Thinakaran
In this study, we present an efficient numerical scheme based on uniform hyperbolic polynomial B-splines for solving the time-fractional diffusion-wave equation involving the Caputo derivative. This equation models various physical phenomena including anomalous diffusion and complex dynamical behavior. The proposed method ensures a smooth and continuous approximation that effectively captures both local and global features of the solution such as sharp gradients and long-range memory effects. The key advantage of uniform hyperbolic polynomial B-splines lies in their flexibility and high accuracy across the computational domain. Stability and convergence analyses are carried out to confirm the method’s robustness and error control. Finally, numerical results are compared with those reported in existing literature to demonstrate the accuracy and reliability of the scheme as process innovation.
{"title":"Solving time fractional diffusion-wave equation using hyperbolic polynomial B-splines: A uniform grid approach","authors":"Aleeza Kiran , Muhammad Yaseen , Aziz Khan , Thabet Abdeljawad , Manar A. Alqudah , Rajermani Thinakaran","doi":"10.1016/j.asej.2025.103868","DOIUrl":"10.1016/j.asej.2025.103868","url":null,"abstract":"<div><div>In this study, we present an efficient numerical scheme based on uniform hyperbolic polynomial B-splines for solving the time-fractional diffusion-wave equation involving the Caputo derivative. This equation models various physical phenomena including anomalous diffusion and complex dynamical behavior. The proposed method ensures a smooth and continuous approximation that effectively captures both local and global features of the solution such as sharp gradients and long-range memory effects. The key advantage of uniform hyperbolic polynomial B-splines lies in their flexibility and high accuracy across the computational domain. Stability and convergence analyses are carried out to confirm the method’s robustness and error control. Finally, numerical results are compared with those reported in existing literature to demonstrate the accuracy and reliability of the scheme as process innovation.</div></div>","PeriodicalId":48648,"journal":{"name":"Ain Shams Engineering Journal","volume":"17 1","pages":"Article 103868"},"PeriodicalIF":5.9,"publicationDate":"2025-11-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145568881","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-21DOI: 10.1016/j.asej.2025.103866
A. Puviarasu , Sudha V K
Federated learning (FL) has been proposed as an effective solution in the context of intrusion detection in IoT networks, where models can be trained collaboratively with the security of raw data protection. In this paper we present a privacy-preserving FL framework based on light weight neural network, differential privacy (DP) and homomorphic encryption (HE). With a dataset of 1,191,264 instances and 47 attributes, the proposed model conducted on the IoT Intrusion Detection Dataset available on Kaggle produces overall accuracy (93.5), precision (94.2), recall (93.4), and the F1-score (94.2), with the detection time of 90–130 ms and no distinction between the attacks, where detection latency was considered in this study. At the attack level the model delivered 94.1 %, 92.5 %, and 93.6 % accuracies on DoS, DDoS, and Mirai respectively, and above 85 % accuracy on Malware and Web-based attacks. DP experiments showed that augmenting the privacy budget parameter 0.5 to 20.0 increased the levels of accuracy by 2.6 % to 94.0 %, and decreased the computational time 150 ms to 121 ms, depicting a compromise between privacy and performance. HE experiments likewise exhibited a negligible accuracy reduction (94.1 % to 93.5 %) between no encryption to complete homomorphic encryption, but required more computation time (120 ms to 200 ms). Devices-level testing demonstrated that the model had > 91 % accuracy at the low-end (0.5 GHz CPU, 128 MB memory) and up to 94.5 % accuracy with 110 ms inference time on powerful processors, irrespective of whether or not the sensor was heterogeneous, demonstrating a viable solution to the heterogeneous IT situation. Audit mechanisms further enhanced greater compliance of 0 % to 99 % with minimal reduction in accuracy (< 0.8 %). The results show that privacy-preserving intrusion detection specifically can be performed with real-time intrusion detection, high detection gene, and privacy guarantees in resource-constrained IoT networks.
{"title":"Enhanced IoT security: privacy-preserving federated learning model for accurate, real-time intrusion detection across devices","authors":"A. Puviarasu , Sudha V K","doi":"10.1016/j.asej.2025.103866","DOIUrl":"10.1016/j.asej.2025.103866","url":null,"abstract":"<div><div>Federated learning (FL) has been proposed as an effective solution in the context of intrusion detection in IoT networks, where models can be trained collaboratively with the security of raw data protection. In this paper we present a privacy-preserving FL framework based on light weight neural network, differential privacy (DP) and homomorphic encryption (HE). With a dataset of 1,191,264 instances and 47 attributes, the proposed model conducted on the IoT Intrusion Detection Dataset available on Kaggle produces overall accuracy (93.5), precision (94.2), recall (93.4), and the F1-score (94.2), with the detection time of 90–130 ms and no distinction between the attacks, where detection latency was considered in this study. At the attack level the model delivered 94.1 %, 92.5 %, and 93.6 % accuracies on DoS, DDoS, and Mirai respectively, and above 85 % accuracy on Malware and Web-based attacks. DP experiments showed that augmenting the privacy budget parameter 0.5 to 20.0 increased the levels of accuracy by 2.6 % to 94.0 %, and decreased the computational time 150 ms to 121 ms, depicting a compromise between privacy and performance. HE experiments likewise exhibited a negligible accuracy reduction (94.1 % to 93.5 %) between no encryption to complete homomorphic encryption, but required more computation time (120 ms to 200 ms). Devices-level testing demonstrated that the model had > 91 % accuracy at the low-end (0.5 GHz CPU, 128 MB memory) and up to 94.5 % accuracy with 110 ms inference time on powerful processors, irrespective of whether or not the sensor was heterogeneous, demonstrating a viable solution to the heterogeneous IT situation. Audit mechanisms further enhanced greater compliance of 0 % to 99 % with minimal reduction in accuracy (< 0.8 %). The results show that privacy-preserving intrusion detection specifically can be performed with real-time intrusion detection, high detection gene, and privacy guarantees in resource-constrained IoT networks.</div></div>","PeriodicalId":48648,"journal":{"name":"Ain Shams Engineering Journal","volume":"17 1","pages":"Article 103866"},"PeriodicalIF":5.9,"publicationDate":"2025-11-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145569349","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this paper, we establish new fractional Hermite–Hadamard type inequalities using -Riemann–Liouville fractional integrals for differentiable -convex functions. By employing -Riemann Riouville fractional integrals and differentiable -convex functions, our results extend and refine the existing inequalities in literature and show the connection between them. We discuss special cases of our derived inequalities which highlight the applicability and novelty of our approach. Furthermore, to support and visualize the theoretical findings, we provide detailed and graphical verifications of the main inequalities. These illustrations offer deeper insights into the behavior of the inequalities and demonstrate their practical relevance. Applications and future research directions are also addressed.
{"title":"Extended fractional hermite-hadamard type integral inequalities for h-convex functions with 2D and 3D graphical illustrations","authors":"Akhtar Abbas , Fazila Fiyaz , Shahid Mubeen , Mdi Begum Jeelani , Ghaliah Alhamzi","doi":"10.1016/j.asej.2025.103819","DOIUrl":"10.1016/j.asej.2025.103819","url":null,"abstract":"<div><div>In this paper, we establish new fractional Hermite–Hadamard type inequalities using <span><math><mi>k</mi></math></span>-Riemann–Liouville fractional integrals for differentiable <span><math><mrow><mi>h</mi></mrow></math></span>-convex functions. By employing <span><math><mi>k</mi></math></span>-Riemann Riouville fractional integrals and differentiable <span><math><mrow><mi>h</mi></mrow></math></span>-convex functions, our results extend and refine the existing inequalities in literature and show the connection between them. We discuss special cases of our derived inequalities which highlight the applicability and novelty of our approach. Furthermore, to support and visualize the theoretical findings, we provide detailed <span><math><mn>2</mn><mi>D</mi></math></span> and <span><math><mn>3</mn><mi>D</mi></math></span> graphical verifications of the main inequalities. These illustrations offer deeper insights into the behavior of the inequalities and demonstrate their practical relevance. Applications and future research directions are also addressed.</div></div>","PeriodicalId":48648,"journal":{"name":"Ain Shams Engineering Journal","volume":"17 1","pages":"Article 103819"},"PeriodicalIF":5.9,"publicationDate":"2025-11-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145569350","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号