Computers & Electrical Engineering最新文献_第10页

Compost maturity prediction and gas emissions monitoring: A sensor-based and interpretable machine learning approach

IF 4 3区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computers & Electrical Engineering

Pub Date : 2025-01-31 DOI: 10.1016/j.compeleceng.2025.110115

Amith Khandakar , Azad Ashraf , Mohamed Arselene Ayari , Amin Esmaeili , Mohannad Aljarrah , Philips Michael , Md. Nahiduzzaman , Hafsa Binte Kibria , Vasiliki Maria Gerokosta , Abdul Ahad Shehbaz , Maryam Abdulla R A Al-Mansoori , Farah Khattab

Here, a sensor-based machine learning approach has been presented to predict the maturity and monitor gas emissions during the composting process. By analyzing key environmental factors and emission data, our study aims to enhance the ecological responsibility of composting as a waste management solution. Our research combines a dedicated sensor system with machine learning. The sensor system, integrated with Arduino Mega 2560 R3 and ESP-32 microcontrollers, wirelessly transmits data for remote monitoring. Meanwhile, our machine learning framework analyzes features such as temperature, C/N ratio, ammonia concentration, pH levels, and nitrate content from ten datasets. After rigorous preprocessing and model training with a robust five-fold cross-validation, we optimize hyperparameters using GridSearchCV. The results highlight that both XGBOOST and CatBOOST excelled in achieving the highest predictive accuracy among the models, each attaining an impressive R² of 0.9912. In particular, XGBOOST demonstrated the lowest mean absolute error (MAE) at 1.1845, while CatBOOST exhibited the lowest mean squared error (MSE) at 1.8382. The interpretability of the model is ensured through LIME and SHAP, making complex models transparent and understandable. The results indicate that the XGBOOST model outperforms the others, achieving the highest predictive accuracy. This groundbreaking approach bridges scientific rigor with practical usability, ensuring responsible waste management for a sustainable future. Real-world applications of our research include more efficient and environmentally friendly waste management systems, reduced environmental impact, and improved compost quality for agricultural use.

{"title":"Compost maturity prediction and gas emissions monitoring: A sensor-based and interpretable machine learning approach","authors":"Amith Khandakar , Azad Ashraf , Mohamed Arselene Ayari , Amin Esmaeili , Mohannad Aljarrah , Philips Michael , Md. Nahiduzzaman , Hafsa Binte Kibria , Vasiliki Maria Gerokosta , Abdul Ahad Shehbaz , Maryam Abdulla R A Al-Mansoori , Farah Khattab","doi":"10.1016/j.compeleceng.2025.110115","DOIUrl":"10.1016/j.compeleceng.2025.110115","url":null,"abstract":"<div><div>Here, a sensor-based machine learning approach has been presented to predict the maturity and monitor gas emissions during the composting process. By analyzing key environmental factors and emission data, our study aims to enhance the ecological responsibility of composting as a waste management solution. Our research combines a dedicated sensor system with machine learning. The sensor system, integrated with Arduino Mega 2560 R3 and ESP-32 microcontrollers, wirelessly transmits data for remote monitoring. Meanwhile, our machine learning framework analyzes features such as temperature, C/N ratio, ammonia concentration, pH levels, and nitrate content from ten datasets. After rigorous preprocessing and model training with a robust five-fold cross-validation, we optimize hyperparameters using GridSearchCV. The results highlight that both XGBOOST and CatBOOST excelled in achieving the highest predictive accuracy among the models, each attaining an impressive R<sup>2</sup> of 0.9912. In particular, XGBOOST demonstrated the lowest mean absolute error (MAE) at 1.1845, while CatBOOST exhibited the lowest mean squared error (MSE) at 1.8382. The interpretability of the model is ensured through LIME and SHAP, making complex models transparent and understandable. The results indicate that the XGBOOST model outperforms the others, achieving the highest predictive accuracy. This groundbreaking approach bridges scientific rigor with practical usability, ensuring responsible waste management for a sustainable future. Real-world applications of our research include more efficient and environmentally friendly waste management systems, reduced environmental impact, and improved compost quality for agricultural use.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110115"},"PeriodicalIF":4.0,"publicationDate":"2025-01-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143148972","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Robust operating strategy for voltage and frequency control in a non-linear hybrid renewable energy-based power system using communication time delay

IF 4 3区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computers & Electrical Engineering

Pub Date : 2025-01-30 DOI: 10.1016/j.compeleceng.2025.110119

Rasmia Irfan , Muhammad Majid Gulzar , Adnan Shakoor , Salman Habib , Hasnain Ahmad , Shahid A. Hasib , Huma Tehreem

Nowadays modern power systems are of interconnected type having both conventional and renewable generation sources. Integration of renewable sources in these modern power systems causes serious stability issues specifically fluctuations of frequency and voltage are one of the major problems. So, to maintain the power quality we need to confront these issues of frequency and voltage fluctuations caused by the intermittent nature of renewable sources such as wind and solar. Addressing these challenges requires advanced control strategies based on real-time monitoring. In this paper, a sine cosine algorithm (SCA) tuned optimal dual mode PI controller with derivative control (DM-PI-DC) is proposed to mitigate frequency and voltage fluctuations. The investigated system comprises two areas having traditional power plants as well as renewable sources while taking into consideration the influence of communication time delays (CTDs). Confrontation of frequency fluctuation is handled by the load frequency control (LFC) loop and regulation of voltage in the power system is accomplished by the automatic voltage regulation (AVR) loop. In order to model a real system, the physical limitations of the power system are also taken into consideration. To manage the power flow, an interline power flow controller (IPFC) is incorporated and to keep the system stable during contingencies redox flow batteries (RFBs) are added to the system. Moreover, to evaluate the competence of the suggested controller it undergoes testing by variable loading, and also the comparison of performance is carried out with the advanced controllers. The detailed analysis showcases that the proposed controller demonstrates an oscillation-free response in 3.3 s whereas other controllers settle in 3.8 s, 6.45 s, 6.2 s, and 3.7 s. Moreover, the proposed controller achieves a 33.33 % improved response, particularly in terms of undershoot. The findings further show that the presented control strategy ensures power quality addressing all the key challenges.

{"title":"Robust operating strategy for voltage and frequency control in a non-linear hybrid renewable energy-based power system using communication time delay","authors":"Rasmia Irfan , Muhammad Majid Gulzar , Adnan Shakoor , Salman Habib , Hasnain Ahmad , Shahid A. Hasib , Huma Tehreem","doi":"10.1016/j.compeleceng.2025.110119","DOIUrl":"10.1016/j.compeleceng.2025.110119","url":null,"abstract":"<div><div>Nowadays modern power systems are of interconnected type having both conventional and renewable generation sources. Integration of renewable sources in these modern power systems causes serious stability issues specifically fluctuations of frequency and voltage are one of the major problems. So, to maintain the power quality we need to confront these issues of frequency and voltage fluctuations caused by the intermittent nature of renewable sources such as wind and solar. Addressing these challenges requires advanced control strategies based on real-time monitoring. In this paper, a sine cosine algorithm (SCA) tuned optimal dual mode PI controller with derivative control (DM-PI-DC) is proposed to mitigate frequency and voltage fluctuations. The investigated system comprises two areas having traditional power plants as well as renewable sources while taking into consideration the influence of communication time delays (CTDs). Confrontation of frequency fluctuation is handled by the load frequency control (LFC) loop and regulation of voltage in the power system is accomplished by the automatic voltage regulation (AVR) loop. In order to model a real system, the physical limitations of the power system are also taken into consideration. To manage the power flow, an interline power flow controller (IPFC) is incorporated and to keep the system stable during contingencies redox flow batteries (RFBs) are added to the system. Moreover, to evaluate the competence of the suggested controller it undergoes testing by variable loading, and also the comparison of performance is carried out with the advanced controllers. The detailed analysis showcases that the proposed controller demonstrates an oscillation-free response in 3.3 s whereas other controllers settle in 3.8 s, 6.45 s, 6.2 s, and 3.7 s. Moreover, the proposed controller achieves a 33.33 % improved response, particularly in terms of undershoot. The findings further show that the presented control strategy ensures power quality addressing all the key challenges.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110119"},"PeriodicalIF":4.0,"publicationDate":"2025-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143148977","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

zk-DASTARK: A quantum-resistant, data authentication and zero-knowledge proof scheme for protecting data feed to smart contracts

IF 4 3区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computers & Electrical Engineering

Pub Date : 2025-01-30 DOI: 10.1016/j.compeleceng.2025.110089

Usama Habib Chaudhry , Razi Arshad , Ayesha Khalid , Indranil Ghosh Ray , Mehdi Hussain

The emergence of blockchain technology and smart contracts revolutionize traditional digital applications such as identity management, supply chain management, banking and financial services with Decentralized Applications (DApps). When DApps are integrated with blockchain technology, blockchain validators can access user-sensitive off-chain data to execute a smart contract. On the one hand, DApps need authentic off-chain input user data to execute a given business scenario properly. On the other hand, users are more concerned about their privacy and are reluctant to share their sensitive data on the blockchain. For instance, healthcare insurance DApp requires sensitive user health data as input. DApp must ensure the privacy and authenticity of the user data given to the smart contract so that blockchain validators can perform operations on the user’s data without disclosing the user’s personal information. However, there is no complete solution to achieve both user privacy and data authenticity at the same time. To address this problem, we propose a highly efficient authenticated zero-knowledge proof scheme named zk-DASTARK by enhancing the standard zk-STARK scheme with a quantum attack-resistant data authentication scheme (CRYSTALS Dilithium digital signature scheme, now FIPS-204 or ML-DSA by the National Institute of Standards and Technology, NIST in the USA). Based on zk-DASTARK, we design a zk-STARKFeed, a zero-knowledge authenticated off-chain data feed mechanism that provides user data privacy and authentication for blockchain-based DApps. The blockchain validators’ computation costs can be significantly reduced using zk-STARKFeed with the integration of the ”compute off-chain and verify on-chain” approach. We have implemented zk-STARKFeed on the IOTA blockchain and performed extensive testing on the healthcare insurance DApp. Our proposed zk-STARKFeed is highly efficient on the IOTA blockchain in such a way that its proof generation takes less than 60 ms (ms) and its proof verification takes less than 10 ms.

{"title":"zk-DASTARK: A quantum-resistant, data authentication and zero-knowledge proof scheme for protecting data feed to smart contracts","authors":"Usama Habib Chaudhry , Razi Arshad , Ayesha Khalid , Indranil Ghosh Ray , Mehdi Hussain","doi":"10.1016/j.compeleceng.2025.110089","DOIUrl":"10.1016/j.compeleceng.2025.110089","url":null,"abstract":"<div><div>The emergence of blockchain technology and smart contracts revolutionize traditional digital applications such as identity management, supply chain management, banking and financial services with Decentralized Applications (DApps). When DApps are integrated with blockchain technology, blockchain validators can access user-sensitive off-chain data to execute a smart contract. On the one hand, DApps need authentic off-chain input user data to execute a given business scenario properly. On the other hand, users are more concerned about their privacy and are reluctant to share their sensitive data on the blockchain. For instance, healthcare insurance DApp requires sensitive user health data as input. DApp must ensure the privacy and authenticity of the user data given to the smart contract so that blockchain validators can perform operations on the user’s data without disclosing the user’s personal information. However, there is no complete solution to achieve both user privacy and data authenticity at the same time. To address this problem, we propose a highly efficient authenticated zero-knowledge proof scheme named zk-DASTARK by enhancing the standard zk-STARK scheme with a quantum attack-resistant data authentication scheme (CRYSTALS Dilithium digital signature scheme, now FIPS-204 or ML-DSA by the National Institute of Standards and Technology, NIST in the USA). Based on zk-DASTARK, we design a zk-STARKFeed, a zero-knowledge authenticated off-chain data feed mechanism that provides user data privacy and authentication for blockchain-based DApps. The blockchain validators’ computation costs can be significantly reduced using zk-STARKFeed with the integration of the ”compute off-chain and verify on-chain” approach. We have implemented zk-STARKFeed on the IOTA blockchain and performed extensive testing on the healthcare insurance DApp. Our proposed zk-STARKFeed is highly efficient on the IOTA blockchain in such a way that its proof generation takes less than 60 ms (ms) and its proof verification takes less than 10 ms.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110089"},"PeriodicalIF":4.0,"publicationDate":"2025-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143148975","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

DBC-MulBiLSTM: A DistilBERT-CNN Feature Fusion Framework enhanced by multi-head self-attention and BiLSTM for smart contract vulnerability detection

IF 4 3区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computers & Electrical Engineering

Pub Date : 2025-01-30 DOI: 10.1016/j.compeleceng.2025.110096

Shujiang Xu , Haochen He , Miodrag J. Mihaljević , Shuhui Zhang , Wei Shao , Qizheng Wang

With the burgeoning of blockchain technology, particularly the Ethereum platform, smart contracts, serving as the core technology of blockchain, have demonstrated immense potential in numerous fields. However, vulnerabilities in smart contracts have also become targets for cyberattacks, potentially leading to significant economic losses. This study introduces a DBC-MulBiLSTM framework designed for the detection of vulnerabilities in smart contracts. The framework first utilizes the lightweight pre-trained model DistilBERT to extract contextual features from smart contracts, while simultaneously utilizing Convolutional Neural Networks (CNN) to identify local features. Through feature fusion, a multi-dimensional feature representation is formed to improve the model’s capabilities to recognize complex vulnerability patterns. Furthermore, the framework incorporates a multi-head self-attention mechanism within the BiLSTM architecture, thereby establishing the MulBiLSTM training framework. This design enables the simultaneous capture of long-range dependencies throughout the entire dataset, enhancing the model’s ability to represent intricate dependencies and contextual information effectively. Experimental results demonstrate that DBC-MulBiLSTM exhibits substantial efficacy in the detection of vulnerabilities within smart contracts, achieving an F1 score of 95.44%, an accuracy rate of 96.57%, and a recall of 95.36%. For various vulnerability types, the model consistently achieves accuracy and F1-scores over 96%, and recall rates above 95%, showcasing efficient and accurate smart contract vulnerability detection capabilities.

{"title":"DBC-MulBiLSTM: A DistilBERT-CNN Feature Fusion Framework enhanced by multi-head self-attention and BiLSTM for smart contract vulnerability detection","authors":"Shujiang Xu , Haochen He , Miodrag J. Mihaljević , Shuhui Zhang , Wei Shao , Qizheng Wang","doi":"10.1016/j.compeleceng.2025.110096","DOIUrl":"10.1016/j.compeleceng.2025.110096","url":null,"abstract":"<div><div>With the burgeoning of blockchain technology, particularly the Ethereum platform, smart contracts, serving as the core technology of blockchain, have demonstrated immense potential in numerous fields. However, vulnerabilities in smart contracts have also become targets for cyberattacks, potentially leading to significant economic losses. This study introduces a DBC-MulBiLSTM framework designed for the detection of vulnerabilities in smart contracts. The framework first utilizes the lightweight pre-trained model DistilBERT to extract contextual features from smart contracts, while simultaneously utilizing Convolutional Neural Networks (CNN) to identify local features. Through feature fusion, a multi-dimensional feature representation is formed to improve the model’s capabilities to recognize complex vulnerability patterns. Furthermore, the framework incorporates a multi-head self-attention mechanism within the BiLSTM architecture, thereby establishing the MulBiLSTM training framework. This design enables the simultaneous capture of long-range dependencies throughout the entire dataset, enhancing the model’s ability to represent intricate dependencies and contextual information effectively. Experimental results demonstrate that DBC-MulBiLSTM exhibits substantial efficacy in the detection of vulnerabilities within smart contracts, achieving an F1 score of 95.44%, an accuracy rate of 96.57%, and a recall of 95.36%. For various vulnerability types, the model consistently achieves accuracy and F1-scores over 96%, and recall rates above 95%, showcasing efficient and accurate smart contract vulnerability detection capabilities.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110096"},"PeriodicalIF":4.0,"publicationDate":"2025-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143148976","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Spatial pyramid attention and affinity inference embedding for unsupervised person re-identification

IF 4 3区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computers & Electrical Engineering

Pub Date : 2025-01-30 DOI: 10.1016/j.compeleceng.2025.110126

Qianyue Duan , Huanjie Tao

Unsupervised person re-identification (Re-ID) aims to learn discriminative features for retrieving person utilizing unlabeled data. Most existing unsupervised person Re-ID methods adopt the generic backbone to extract features for clustering to generate pseudo labels and utilize the pseudo labels to train the model. However, due to the lack of accurate category supervision, the generic backbone inevitably extracts interfering features, which degrade the quality of pseudo-labels. Besides, many methods only utilize the similarity between query and gallery images for matching person and ignore the use of affinity information between gallery images. To solve the above issues, we propose a spatial pyramid attention and affinity inference embedding network for unsupervised person Re-ID. We explore the benefit of attention mechanisms in unsupervised person Re-ID, where research is currently limited. We adopt the spatial pyramid attention (SPA) to aggregate structural information at different scales and ensures enough utilization of structural information during attention learning. With the help of SPA, the model reduces the extraction of interfering features, ensuring that it can learn more discriminative for clustering to improve pseudo-label quality. In addition, the affinity inference module (AIM) is utilized to optimize the distance between the query images and the gallery images by additionally using affinity information between gallery images. Extensive experiments on three datasets demonstrate that our method achieves competitive performance. Especially, our method achieves Rank-1 accuracy of 77.1 % on the MSMT17 dataset, outperforming the recent unsupervised work DCMIP by 7+%. Our code will be released at: https://github.com/wanderer1230/SPAENet.

{"title":"Spatial pyramid attention and affinity inference embedding for unsupervised person re-identification","authors":"Qianyue Duan , Huanjie Tao","doi":"10.1016/j.compeleceng.2025.110126","DOIUrl":"10.1016/j.compeleceng.2025.110126","url":null,"abstract":"<div><div>Unsupervised person re-identification (<em>Re</em>-ID) aims to learn discriminative features for retrieving person utilizing unlabeled data. Most existing unsupervised person <em>Re</em>-ID methods adopt the generic backbone to extract features for clustering to generate pseudo labels and utilize the pseudo labels to train the model. However, due to the lack of accurate category supervision, the generic backbone inevitably extracts interfering features, which degrade the quality of pseudo-labels. Besides, many methods only utilize the similarity between query and gallery images for matching person and ignore the use of affinity information between gallery images. To solve the above issues, we propose a spatial pyramid attention and affinity inference embedding network for unsupervised person <em>Re</em>-ID. We explore the benefit of attention mechanisms in unsupervised person <em>Re</em>-ID, where research is currently limited. We adopt the spatial pyramid attention (SPA) to aggregate structural information at different scales and ensures enough utilization of structural information during attention learning. With the help of SPA, the model reduces the extraction of interfering features, ensuring that it can learn more discriminative for clustering to improve pseudo-label quality. In addition, the affinity inference module (AIM) is utilized to optimize the distance between the query images and the gallery images by additionally using affinity information between gallery images. Extensive experiments on three datasets demonstrate that our method achieves competitive performance. Especially, our method achieves Rank-1 accuracy of 77.1 % on the MSMT17 dataset, outperforming the recent unsupervised work DCMIP by 7+%. Our code will be released at: <span><span>https://github.com/wanderer1230/SPAENet</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110126"},"PeriodicalIF":4.0,"publicationDate":"2025-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143097541","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Navigating cybersecurity training: A comprehensive review

IF 4 3区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computers & Electrical Engineering

Pub Date : 2025-01-29 DOI: 10.1016/j.compeleceng.2025.110097

Saif Al-Dean Qawasmeh , Ali Abdullah S. AlQahtani , Muhammad Khurram Khan

In the dynamic realm of cybersecurity, awareness training is crucial for strengthening defenses against cyber threats. This survey examines a spectrum of cybersecurity awareness training methods, analyzing traditional, technology-based, and innovative strategies. It evaluates the principles, efficacy, and constraints of each method, presenting a comparative analysis that highlights their pros and cons. The study also investigates emerging trends like artificial intelligence and extended reality, discussing their prospective influence on the future of cybersecurity training. Additionally, it addresses implementation challenges and proposes solutions, drawing on insights from real-world case studies. The goal is to bolster the understanding of cybersecurity awareness training’s current landscape, offering valuable perspectives for both practitioners and scholars.

引用次数: 0

MPPT controller improvement for a PEM fuel cell system based on Gaussian Process Regression with a digital twin

IF 4 3区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computers & Electrical Engineering

Pub Date : 2025-01-29 DOI: 10.1016/j.compeleceng.2025.110101

Jokin Uralde , Oscar Barambones , Jesus Sanchez , Isidro Calvo , Asier del Rio

Hydrogen, due to its high energy density, stands out as an energy storage method for renewable energies in order to reduce the impact of the energy sector on global warming. Proton Exchange Membrane Fuel Cells (PEMFC) are responsible for converting the stored hydrogen into electrical energy and in order to obtain the highest energy conversion efficiency, the maximum power point (MPP) of the voltage-power curve of the fuel cell must be reached. Traditional Maximum Power Point Tracking (MPPT) algorithms, such as Perturb and Observe (P&O) or controllers such us Proportional Integral Derivative (PID) controller, are easy to implement, but must strike a balance between fast response and accurate control. Other more complex controllers such as Fuzzy Logic Control (FLC) or neural networks achieve better performance but at a higher computational cost. This paper presents a combination of a conventional Sliding Mode Control (SMC) and a machine learning Gaussian Process Regression (GPR) algorithm that provides a reference duty cycle reaching a point close to the MPP which is then used by the SMC to obtain the actual MPP. A Digital Twin of the PEMFC and a DC/DC converter, which allow a fast and large data-set generation, are used for the generation of the GPR algorithm. The proposed control is compared with a conventional SMC and performance improvements are observed using the Integral of the Absolute Error (IAE) metric. The results show, in a control initiation test, a 67% improvement in the IAE metric of the proposed control over the conventional SMC. In a load change test, the proposed control also outperforms the conventional SMC by 42.9%.

{"title":"MPPT controller improvement for a PEM fuel cell system based on Gaussian Process Regression with a digital twin","authors":"Jokin Uralde , Oscar Barambones , Jesus Sanchez , Isidro Calvo , Asier del Rio","doi":"10.1016/j.compeleceng.2025.110101","DOIUrl":"10.1016/j.compeleceng.2025.110101","url":null,"abstract":"<div><div>Hydrogen, due to its high energy density, stands out as an energy storage method for renewable energies in order to reduce the impact of the energy sector on global warming. Proton Exchange Membrane Fuel Cells (PEMFC) are responsible for converting the stored hydrogen into electrical energy and in order to obtain the highest energy conversion efficiency, the maximum power point (MPP) of the voltage-power curve of the fuel cell must be reached. Traditional Maximum Power Point Tracking (MPPT) algorithms, such as Perturb and Observe (P&O) or controllers such us Proportional Integral Derivative (PID) controller, are easy to implement, but must strike a balance between fast response and accurate control. Other more complex controllers such as Fuzzy Logic Control (FLC) or neural networks achieve better performance but at a higher computational cost. This paper presents a combination of a conventional Sliding Mode Control (SMC) and a machine learning Gaussian Process Regression (GPR) algorithm that provides a reference duty cycle reaching a point close to the MPP which is then used by the SMC to obtain the actual MPP. A Digital Twin of the PEMFC and a DC/DC converter, which allow a fast and large data-set generation, are used for the generation of the GPR algorithm. The proposed control is compared with a conventional SMC and performance improvements are observed using the Integral of the Absolute Error (IAE) metric. The results show, in a control initiation test, a 67% improvement in the IAE metric of the proposed control over the conventional SMC. In a load change test, the proposed control also outperforms the conventional SMC by 42.9%.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110101"},"PeriodicalIF":4.0,"publicationDate":"2025-01-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143149961","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Proactive and data-centric Internet of Things-based fog computing architecture for effective policing in smart cities

IF 4 3区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computers & Electrical Engineering

Pub Date : 2025-01-29 DOI: 10.1016/j.compeleceng.2024.110030

Ateeq Ur Rehman Butt , Tanzila Saba , Inayat Khan , Tariq Mahmood , Amjad Rehman Khan , Sushil Kumar Singh , Yousef Ibrahim Daradkeh , Inam Ullah

Smart surveillance is crucial for improving citizen security and ensuring a sustainable environment for routine tasks, particularly within intelligent transportation systems (ITS). However, it can be costly and burden taxpayers. The lack of public interaction makes it difficult for police to arrest and conduct investigations. Additionally, incidents increase due to similar patterns, making smart surveillance essential for reporting and addressing these issues. Smart devices such as sensors or actuators installed on the roads and within vehicles are critical components of any smart surveillance and ITS framework. This integration enhances system agility and facilitates proactive rather than reactive responses. It empowers security agencies to plan more effectively and respond swiftly during emergencies. The incorporation of cloud computing capabilities transforms traditional surveillance and ITS operations. Employing the Internet of Things (IoT) with edge or cloud computing extensions, such as fog computing, modernizes the management of security gadgets for Field Forces. This study investigates a smart surveillance fog-enabled approach to reduce response times for aiding agencies within ITS. By optimizing individual journeys through an RFID-based passing system, incidents are reported promptly to the nearest field force, enhancing overall ITS efficiency. The proactive approach improves resource consumption (energy, CPU, and network usage) compared to traditional reactive methods. The fog-enabled experiments demonstrated a CPU efficiency of approximately 95.76%, significantly outperforming the Cloud-only deployment, achieving a maximum average efficiency of 92.12%. Experimental evaluations in a simulation environment demonstrate that the proposed method significantly outperforms conventional approaches, marking a substantial advancement in IoT-aided ITS.

{"title":"Proactive and data-centric Internet of Things-based fog computing architecture for effective policing in smart cities","authors":"Ateeq Ur Rehman Butt , Tanzila Saba , Inayat Khan , Tariq Mahmood , Amjad Rehman Khan , Sushil Kumar Singh , Yousef Ibrahim Daradkeh , Inam Ullah","doi":"10.1016/j.compeleceng.2024.110030","DOIUrl":"10.1016/j.compeleceng.2024.110030","url":null,"abstract":"<div><div>Smart surveillance is crucial for improving citizen security and ensuring a sustainable environment for routine tasks, particularly within intelligent transportation systems (ITS). However, it can be costly and burden taxpayers. The lack of public interaction makes it difficult for police to arrest and conduct investigations. Additionally, incidents increase due to similar patterns, making smart surveillance essential for reporting and addressing these issues. Smart devices such as sensors or actuators installed on the roads and within vehicles are critical components of any smart surveillance and ITS framework. This integration enhances system agility and facilitates proactive rather than reactive responses. It empowers security agencies to plan more effectively and respond swiftly during emergencies. The incorporation of cloud computing capabilities transforms traditional surveillance and ITS operations. Employing the Internet of Things (IoT) with edge or cloud computing extensions, such as fog computing, modernizes the management of security gadgets for Field Forces. This study investigates a smart surveillance fog-enabled approach to reduce response times for aiding agencies within ITS. By optimizing individual journeys through an RFID-based passing system, incidents are reported promptly to the nearest field force, enhancing overall ITS efficiency. The proactive approach improves resource consumption (energy, CPU, and network usage) compared to traditional reactive methods. The fog-enabled experiments demonstrated a CPU efficiency of approximately 95.76%, significantly outperforming the Cloud-only deployment, achieving a maximum average efficiency of 92.12%. Experimental evaluations in a simulation environment demonstrate that the proposed method significantly outperforms conventional approaches, marking a substantial advancement in IoT-aided ITS.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110030"},"PeriodicalIF":4.0,"publicationDate":"2025-01-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143097548","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Secure and reversible fragile watermarking for accurate authentication and tamper localization in medical images

IF 4 3区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computers & Electrical Engineering

Pub Date : 2025-01-29 DOI: 10.1016/j.compeleceng.2025.110072

Riadh Bouarroudj , Feryel Souami , Fatma Zohra Bellala , Nabil Zerrouki , Fouzi Harrou , Ying Sun

As digital image sharing becomes more prevalent in healthcare, ensuring image security without compromising diagnostic quality is crucial. Reversible watermarking provides an effective solution by enabling authentication and complete restoration of the original image. This study presents a fully blind and reversible fragile watermarking method for authenticating color and grayscale medical images while precisely localizing tampered regions. The proposed approach generates the watermark directly from the host image using a 2-level Discrete Wavelet Transform (DWT) and encrypts it with logistic mapping for enhanced security, eliminating the need for separate storage. The cover image is divided into 4 × 4 blocks, and the Discrete Fourier Transform (DFT) is applied to each block. High-frequency coefficients are modified during embedding to incorporate the watermark, while the extraction process accurately retrieves and decrypts it to detect and localize tampered areas. This method ensures that the original image can be fully restored if no tampering is detected, offering significant advancements in image authentication, tamper detection, and image restoration for sensitive medical applications. Experimental results across seven different datasets demonstrate that the method achieves high-quality watermarked images with a Peak Signal-to-Noise Ratio (PSNR) exceeding 88 dB, and high watermark extraction accuracy, while maintaining a payload of 0.5 bits per pixel (BPP). It also shows high sensitivity to multiple attacks, accurately localizing tampered areas as small as 4 × 4 pixels, or 0.005% of the image size, which surpasses the accuracy achieved by other models in the literature.

{"title":"Secure and reversible fragile watermarking for accurate authentication and tamper localization in medical images","authors":"Riadh Bouarroudj , Feryel Souami , Fatma Zohra Bellala , Nabil Zerrouki , Fouzi Harrou , Ying Sun","doi":"10.1016/j.compeleceng.2025.110072","DOIUrl":"10.1016/j.compeleceng.2025.110072","url":null,"abstract":"<div><div>As digital image sharing becomes more prevalent in healthcare, ensuring image security without compromising diagnostic quality is crucial. Reversible watermarking provides an effective solution by enabling authentication and complete restoration of the original image. This study presents a fully blind and reversible fragile watermarking method for authenticating color and grayscale medical images while precisely localizing tampered regions. The proposed approach generates the watermark directly from the host image using a 2-level Discrete Wavelet Transform (DWT) and encrypts it with logistic mapping for enhanced security, eliminating the need for separate storage. The cover image is divided into 4 × 4 blocks, and the Discrete Fourier Transform (DFT) is applied to each block. High-frequency coefficients are modified during embedding to incorporate the watermark, while the extraction process accurately retrieves and decrypts it to detect and localize tampered areas. This method ensures that the original image can be fully restored if no tampering is detected, offering significant advancements in image authentication, tamper detection, and image restoration for sensitive medical applications. Experimental results across seven different datasets demonstrate that the method achieves high-quality watermarked images with a Peak Signal-to-Noise Ratio (PSNR) exceeding 88 dB, and high watermark extraction accuracy, while maintaining a payload of 0.5 bits per pixel (BPP). It also shows high sensitivity to multiple attacks, accurately localizing tampered areas as small as 4 × 4 pixels, or 0.005% of the image size, which surpasses the accuracy achieved by other models in the literature.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110072"},"PeriodicalIF":4.0,"publicationDate":"2025-01-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143149959","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Comprehensive attention transformer for multi-label segmentation of medical images based on multi-scale feature fusion

IF 4 3区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computers & Electrical Engineering

Pub Date : 2025-01-28 DOI: 10.1016/j.compeleceng.2025.110100

Hangyuan Cheng , Xiaoxin Guo , Guangqi Yang , Cong Chen , Hongliang Dong

Transformer-based models often neglect the convolutional networks’ capability to extract local features. Most U-shaped models only utilize the multi-scale features from the encoder’s output and focus solely on the final layer of the decoder’s output. Moreover, most typical skip connections can be configured only between the encoder and the decoder in the same layer without any potential optimization An innovative comprehensive attention Transformer (CAFormer) is proposed to address the issue of long-range relations and local features in multi-label segmentation of medical images, which adopts a U-shaped hierarchical encoder–decoder structure. A mixed-attention Transformer (MATrans) module is devised to extract multi-scale features from the encoders and establish multiple encoder–decoder connections using channel-wise cross-attention and self-attention, which can automatically configure the optimal skip connections. During the upsampling, a channel-based feature fusion module is proposed to focus on the important channel-based features. A comprehensive attention module (CAM) is designed to extract global context and local features by integrating an enhanced Transformer module, a channel and spatial attention module. Additionally, the encoder’s multi-scale features undergo the hierarchical prediction link through the proposed multi-scale aggregation module (MSAM) for the final prediction rather than directly using the output of the last layer of the decoder as the segmentation outcome. The experiments show that the CAFormer is efficient, robust, achieves the DSC of 82.26 and the HD of 15.26 on the Synapse dataset, and outperforms other state-of-the-art models. The code and pre-trained models are available at https://github.com/zed-kingc/CAFormer.

{"title":"Comprehensive attention transformer for multi-label segmentation of medical images based on multi-scale feature fusion","authors":"Hangyuan Cheng , Xiaoxin Guo , Guangqi Yang , Cong Chen , Hongliang Dong","doi":"10.1016/j.compeleceng.2025.110100","DOIUrl":"10.1016/j.compeleceng.2025.110100","url":null,"abstract":"<div><div>Transformer-based models often neglect the convolutional networks’ capability to extract local features. Most U-shaped models only utilize the multi-scale features from the encoder’s output and focus solely on the final layer of the decoder’s output. Moreover, most typical skip connections can be configured only between the encoder and the decoder in the same layer without any potential optimization An innovative comprehensive attention Transformer (CAFormer) is proposed to address the issue of long-range relations and local features in multi-label segmentation of medical images, which adopts a U-shaped hierarchical encoder–decoder structure. A mixed-attention Transformer (MATrans) module is devised to extract multi-scale features from the encoders and establish multiple encoder–decoder connections using channel-wise cross-attention and self-attention, which can automatically configure the optimal skip connections. During the upsampling, a channel-based feature fusion module is proposed to focus on the important channel-based features. A comprehensive attention module (CAM) is designed to extract global context and local features by integrating an enhanced Transformer module, a channel and spatial attention module. Additionally, the encoder’s multi-scale features undergo the hierarchical prediction link through the proposed multi-scale aggregation module (MSAM) for the final prediction rather than directly using the output of the last layer of the decoder as the segmentation outcome. The experiments show that the CAFormer is efficient, robust, achieves the DSC of 82.26 and the HD of 15.26 on the Synapse dataset, and outperforms other state-of-the-art models. The code and pre-trained models are available at <span><span>https://github.com/zed-kingc/CAFormer</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110100"},"PeriodicalIF":4.0,"publicationDate":"2025-01-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143149960","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0