Pub Date : 2025-02-19DOI: 10.1109/ACCESS.2025.3543599
Pan Wang;Toru Kurihara;Jun Yu
Motion deblurring in scenes involving small moving objects or low-illumination conditions is challenging. This paper presents an effective deep-learning solution that utilizes correlation images as key auxiliaries to address the problem. The correlation image, produced by a three-phase correlation image sensor (3PCIS), is a temporal correlation between incident light and reference signals within a frame time, which encodes intensity changes of incident light over the exposure time. Since correlation images explicitly record motion information lost during the blurring process during exposure, they can be used for accurately identifying the location and degree of blur, especially in low-illumination conditions and scenarios with small moving objects. Therefore, we combine correlation images and motion-blurred images as inputs and build a two-stream network for motion deblurring. Two key designs in our model are 1) Shared-gated Block (SGB), which enables information exchange between the two encoders and selectively allows useful information to pass through the network to obtain high-quality output; 2) a Motion-guided Block (MGB), decoding process that can draw more attention to the blurred areas in the image, thereby achieving clearer textures and details restoration in the blurred areas. The experimental results show that our model not only can successfully eliminate the motion blur in the above challenging scenarios, but also achieves a state-of-the-art 36.02dB in Peak Signal-to-Noise Ratio (PSNR) on the GoPro dataset with simulated correlation images.
{"title":"CorNet: Enhancing Motion Deblurring in Challenging Scenarios Using Correlation Image Sensor","authors":"Pan Wang;Toru Kurihara;Jun Yu","doi":"10.1109/ACCESS.2025.3543599","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3543599","url":null,"abstract":"Motion deblurring in scenes involving small moving objects or low-illumination conditions is challenging. This paper presents an effective deep-learning solution that utilizes correlation images as key auxiliaries to address the problem. The correlation image, produced by a three-phase correlation image sensor (3PCIS), is a temporal correlation between incident light and reference signals within a frame time, which encodes intensity changes of incident light over the exposure time. Since correlation images explicitly record motion information lost during the blurring process during exposure, they can be used for accurately identifying the location and degree of blur, especially in low-illumination conditions and scenarios with small moving objects. Therefore, we combine correlation images and motion-blurred images as inputs and build a two-stream network for motion deblurring. Two key designs in our model are 1) Shared-gated Block (SGB), which enables information exchange between the two encoders and selectively allows useful information to pass through the network to obtain high-quality output; 2) a Motion-guided Block (MGB), decoding process that can draw more attention to the blurred areas in the image, thereby achieving clearer textures and details restoration in the blurred areas. The experimental results show that our model not only can successfully eliminate the motion blur in the above challenging scenarios, but also achieves a state-of-the-art 36.02dB in Peak Signal-to-Noise Ratio (PSNR) on the GoPro dataset with simulated correlation images.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"13 ","pages":"33834-33848"},"PeriodicalIF":3.4,"publicationDate":"2025-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10892137","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143512875","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-19DOI: 10.1109/ACCESS.2025.3543751
Shahid Tufail;Hasan Iqbal;Mohd Tariq;Arif I. Sarwat
Cyberattacks, especially data injection attacks, are becoming more common as smart grids are increasingly interconnected. In addition, accurate and unbiased high-quality data is required for model training. Most of the data we collect from the real world is sparse, incomplete, inconsistent, and skewed. To address these issues, we have proposed a framework to detect such attacks in this study. Using a stacked autoencoder architecture, synthetic instances of minority class data were generated. The generated classes address the imbalances in the data to enhance the generalizability of the model and address diverse attack scenarios. Various machine learning algorithms were evaluated, and the Random Forest (RF) model consistently achieved superior accuracy, ranging from 99.32% to 95.89%. In particular, traditional algorithms such as Logistic Regression (LR) exhibited sensitivity to dimensionality reductions, experiencing a 16.96% accuracy drop when the principal components were reduced from all to 10. In contrast, RF demonstrated resilience, with only a 1.67% mean accuracy drop under similar conditions. Both RF and XGBoost (XGB) emerged as standout models, showcasing high accuracy and robust performance even with dimensionality reduction via principal component analysis (PCA). However, reducing PCA components from 10 to 5 led to performance decreases in all models. The Support Vector Machine (SVM) Classifier shows the highest accuracy drop of 14.21%. This study shows the importance of understanding algorithmic behavior and data features and how it can impact the performance of ML models. This analysis will strengthen cybersecurity in smart grids and focusing on the critical need for careful feature selection and tuning, particularly for models sensitive to dimensionality reduction.
{"title":"A Hybrid Machine Learning-Based Framework for Data Injection Attack Detection in Smart Grids Using PCA and Stacked Autoencoders","authors":"Shahid Tufail;Hasan Iqbal;Mohd Tariq;Arif I. Sarwat","doi":"10.1109/ACCESS.2025.3543751","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3543751","url":null,"abstract":"Cyberattacks, especially data injection attacks, are becoming more common as smart grids are increasingly interconnected. In addition, accurate and unbiased high-quality data is required for model training. Most of the data we collect from the real world is sparse, incomplete, inconsistent, and skewed. To address these issues, we have proposed a framework to detect such attacks in this study. Using a stacked autoencoder architecture, synthetic instances of minority class data were generated. The generated classes address the imbalances in the data to enhance the generalizability of the model and address diverse attack scenarios. Various machine learning algorithms were evaluated, and the Random Forest (RF) model consistently achieved superior accuracy, ranging from 99.32% to 95.89%. In particular, traditional algorithms such as Logistic Regression (LR) exhibited sensitivity to dimensionality reductions, experiencing a 16.96% accuracy drop when the principal components were reduced from all to 10. In contrast, RF demonstrated resilience, with only a 1.67% mean accuracy drop under similar conditions. Both RF and XGBoost (XGB) emerged as standout models, showcasing high accuracy and robust performance even with dimensionality reduction via principal component analysis (PCA). However, reducing PCA components from 10 to 5 led to performance decreases in all models. The Support Vector Machine (SVM) Classifier shows the highest accuracy drop of 14.21%. This study shows the importance of understanding algorithmic behavior and data features and how it can impact the performance of ML models. This analysis will strengthen cybersecurity in smart grids and focusing on the critical need for careful feature selection and tuning, particularly for models sensitive to dimensionality reduction.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"13 ","pages":"33783-33798"},"PeriodicalIF":3.4,"publicationDate":"2025-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10892133","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143512894","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-19DOI: 10.1109/ACCESS.2025.3543659
David A. Zambrano-Prada;Abdelali El Aroudi;Oswaldo López-Santos;Luis Vázquez-Seisdedos;Luis Martí-Nez-Salamero
In this paper, output voltage regulation in a boost converter with constant power load (CPL) is carried out by means of sliding-mode control (SMC) with an estimation loop of the output power. The estimation procedure is based on the integral of an odd-symmetric function of the output voltage error, which confers an adaptive nature to the switching regulator. Rational, trigonometric, sigmoid and $mathop {mathrm {sign}}nolimits $ -type odd-symmetric functions are analyzed to select the best candidate for output power estimation. In addition, a two-degree polynomial surface is considered to induce the sliding motions. Subsequently, the corresponding conditions for the existence of the sliding mode and for the stability of the equilibrium point including the estimation dynamics are derived. One of the main features of this proposal is that the resulting controller can be implemented analogically, requiring operational amplifier-based circuits plus a divider. PSIM$^{unicode {0x00A9}} $ and MATLAB$^{unicode {0x00A9}} $ simulations show a fast recovery in response to large-signal disturbances in the load power and zero steady-state output voltage error. Experimental results obtained from a 500 W prototype are in perfect agreement with both theoretical predictions and numerical simulations.
{"title":"Adaptive Sliding Mode Control of a Boost Converter With Unknown Constant Power Load","authors":"David A. Zambrano-Prada;Abdelali El Aroudi;Oswaldo López-Santos;Luis Vázquez-Seisdedos;Luis Martí-Nez-Salamero","doi":"10.1109/ACCESS.2025.3543659","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3543659","url":null,"abstract":"In this paper, output voltage regulation in a boost converter with constant power load (CPL) is carried out by means of sliding-mode control (SMC) with an estimation loop of the output power. The estimation procedure is based on the integral of an odd-symmetric function of the output voltage error, which confers an adaptive nature to the switching regulator. Rational, trigonometric, sigmoid and <inline-formula> <tex-math>$mathop {mathrm {sign}}nolimits $ </tex-math></inline-formula>-type odd-symmetric functions are analyzed to select the best candidate for output power estimation. In addition, a two-degree polynomial surface is considered to induce the sliding motions. Subsequently, the corresponding conditions for the existence of the sliding mode and for the stability of the equilibrium point including the estimation dynamics are derived. One of the main features of this proposal is that the resulting controller can be implemented analogically, requiring operational amplifier-based circuits plus a divider. PSIM<inline-formula> <tex-math>$^{unicode {0x00A9}} $ </tex-math></inline-formula> and MATLAB<inline-formula> <tex-math>$^{unicode {0x00A9}} $ </tex-math></inline-formula> simulations show a fast recovery in response to large-signal disturbances in the load power and zero steady-state output voltage error. Experimental results obtained from a 500 W prototype are in perfect agreement with both theoretical predictions and numerical simulations.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"13 ","pages":"33714-33732"},"PeriodicalIF":3.4,"publicationDate":"2025-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10892111","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143512834","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper proposes a method for in-memory true random number generation (TRNG) by leveraging the dual functionality of spin-orbit torque based magnetic tunnel junction (SOT-MTJ) while showcasing its efficacy in hardware-efficient Grain stream ciphers for lightweight cryptographic applications. Depending upon its mode of operation, SOT-MTJ acts as both a memory element and a true random number generator. To demonstrate its practical application, SOT-MTJ based non-volatile flip flop (NVFF) is designed which is further utilized to implement Grain-128 stream cipher, as a case study. The SOT-MTJ based NVFF not only carries out the standard shift operation for cipher implementation but also functions as an in-situ initial vector generator for generating key stream, eliminating the need for an additional TRNG circuit. The results show that the proposed Grain-128 cipher design is $5.6times $ and $2.5times $ more energy efficient and $5times $ and $2times $ faster as compared to STT and SOT-MTJ based designs. Furthermore, in comparison to CMOS based cipher design, the proposed technique shows nearly $sim 34times $ more efficiency in terms of area overhead. The proposed approach holds huge promise for resource-constrained cryptographic applications in edge devices.
{"title":"SOT-MTJ-Based Non-Volatile Flip-Flop With In-Memory Randomness for Application in Grain Stream Ciphers","authors":"Arshid Nisar;Furqan Zahoor;Sidhaant Sachin Thakker;Kunal Kranti Das;Subhamoy Maitra;Brajesh Kumar Kaushik;Anupam Chattopadhyay","doi":"10.1109/ACCESS.2025.3543733","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3543733","url":null,"abstract":"This paper proposes a method for in-memory true random number generation (TRNG) by leveraging the dual functionality of spin-orbit torque based magnetic tunnel junction (SOT-MTJ) while showcasing its efficacy in hardware-efficient Grain stream ciphers for lightweight cryptographic applications. Depending upon its mode of operation, SOT-MTJ acts as both a memory element and a true random number generator. To demonstrate its practical application, SOT-MTJ based non-volatile flip flop (NVFF) is designed which is further utilized to implement Grain-128 stream cipher, as a case study. The SOT-MTJ based NVFF not only carries out the standard shift operation for cipher implementation but also functions as an in-situ initial vector generator for generating key stream, eliminating the need for an additional TRNG circuit. The results show that the proposed Grain-128 cipher design is <inline-formula> <tex-math>$5.6times $ </tex-math></inline-formula> and <inline-formula> <tex-math>$2.5times $ </tex-math></inline-formula> more energy efficient and <inline-formula> <tex-math>$5times $ </tex-math></inline-formula> and <inline-formula> <tex-math>$2times $ </tex-math></inline-formula> faster as compared to STT and SOT-MTJ based designs. Furthermore, in comparison to CMOS based cipher design, the proposed technique shows nearly <inline-formula> <tex-math>$sim 34times $ </tex-math></inline-formula> more efficiency in terms of area overhead. The proposed approach holds huge promise for resource-constrained cryptographic applications in edge devices.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"13 ","pages":"34677-34686"},"PeriodicalIF":3.4,"publicationDate":"2025-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10892132","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143513011","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-19DOI: 10.1109/ACCESS.2025.3543748
Gonzalo Martinez Medina;Krystel K. Castillo-Villar;Omar Abbaas
This paper addresses a critical challenge that utility providers face as commercial electric vehicle (EV) fleets rapidly expand. Specifically, it focuses on optimizing charging infrastructure for medium- and heavy-duty electric vehicles while managing constrained grid capacity. Businesses’ increasing adoption of mid-size to heavy EV fleets has created a significant surge in electricity demand, often exceeding the local grid’s ability to support charging at vehicles’ base locations. Supply chain constraints that hinder timely infrastructure upgrades exacerbate this mismatch between demand and capacity. We present an optimization model for EV fleet charging location assignment that tackles this issue. Our approach considers multiple commercial fleet operators, each with a set of base locations for their vehicles. The model accounts for limited charging capacity at these bases and proposes strategically placing charging hubs in areas with excess grid capacity. We incorporate a flexible incentive framework into our model to encourage the use of these hubs and other non-base charging locations. The primary objective of this study is to optimize the allocation of charging resources for commercial EV fleets and to maintain grid stability in the face of rapidly growing demand. Our model integrates fleet operational constraints, grid limitations, and incentive structures to provide a comprehensive solution that benefits fleet operators and utility providers. To validate our approach, we perform a series of computational experiments based on realistic data from the city of San Antonio, TX, a major urban center in Texas. These simulations demonstrate the model’s effectiveness in managing peak demand, optimizing resource utilization, and providing actionable insights for infrastructure planning.
{"title":"Optimization Model for Electric Vehicle (EV) Fleet Charging Location Assignment","authors":"Gonzalo Martinez Medina;Krystel K. Castillo-Villar;Omar Abbaas","doi":"10.1109/ACCESS.2025.3543748","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3543748","url":null,"abstract":"This paper addresses a critical challenge that utility providers face as commercial electric vehicle (EV) fleets rapidly expand. Specifically, it focuses on optimizing charging infrastructure for medium- and heavy-duty electric vehicles while managing constrained grid capacity. Businesses’ increasing adoption of mid-size to heavy EV fleets has created a significant surge in electricity demand, often exceeding the local grid’s ability to support charging at vehicles’ base locations. Supply chain constraints that hinder timely infrastructure upgrades exacerbate this mismatch between demand and capacity. We present an optimization model for EV fleet charging location assignment that tackles this issue. Our approach considers multiple commercial fleet operators, each with a set of base locations for their vehicles. The model accounts for limited charging capacity at these bases and proposes strategically placing charging hubs in areas with excess grid capacity. We incorporate a flexible incentive framework into our model to encourage the use of these hubs and other non-base charging locations. The primary objective of this study is to optimize the allocation of charging resources for commercial EV fleets and to maintain grid stability in the face of rapidly growing demand. Our model integrates fleet operational constraints, grid limitations, and incentive structures to provide a comprehensive solution that benefits fleet operators and utility providers. To validate our approach, we perform a series of computational experiments based on realistic data from the city of San Antonio, TX, a major urban center in Texas. These simulations demonstrate the model’s effectiveness in managing peak demand, optimizing resource utilization, and providing actionable insights for infrastructure planning.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"13 ","pages":"34160-34176"},"PeriodicalIF":3.4,"publicationDate":"2025-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10892131","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143513024","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Realistic Global Navigation Satellite System (GNSS) synthetic data is essential for the research and development of vehicular applications, such as Advanced Driver Assistance Systems (ADAS), autonomous driving, and solutions or scenarios that are difficult and expensive to test in the real world, such as vehicular cooperative positioning. However, generating GNSS synthetic data is complex due to satellite dynamics, signal characteristics, and various noise and error sources. This complexity increases in automotive contexts by vehicle movement and environmental factors influencing signal propagation, with multipath effects being particularly challenging to simulate accurately. This paper introduces a novel pipeline that leverages a 3D virtual environment to produce more realistic GNSS synthetic data for automotive applications. The pipeline integrates the CARLA Simulator and GPSoft’s SatNav Toolbox for Matlab, with custom-developed modules that generate raw GNSS measurements incorporating environment- and location-specific multipath effects. Our contributions include a tailored simulation pipeline for automotive applications, with integration of GNSS satellite orbits within CARLA, a dynamic multipath model reflecting obstacles in the simulated environment, and a synthetic dataset generated by this approach available to the community. Evaluation on CARLA’s Town03 map showed that while standard multipath models result in unrealistic uniform effects, our dynamic model produces effects that correlate with the vehicle’s surroundings, accurately reflecting real-world conditions such as increased errors in urban areas and lack of signals in tunnels. This approach can support the research, development, and validation of GNSS positioning algorithms and Artificial Intelligence (AI) model training, with potential applications extending also beyond the automotive context.
{"title":"GNSS Simulation for Automotive: Introducing 3D Scene-Dependent Multipath With CARLA","authors":"Cristiano Pendão;Ivo Silva;Fabricio Botelho;Hélder Silva","doi":"10.1109/ACCESS.2025.3543752","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3543752","url":null,"abstract":"Realistic Global Navigation Satellite System (GNSS) synthetic data is essential for the research and development of vehicular applications, such as Advanced Driver Assistance Systems (ADAS), autonomous driving, and solutions or scenarios that are difficult and expensive to test in the real world, such as vehicular cooperative positioning. However, generating GNSS synthetic data is complex due to satellite dynamics, signal characteristics, and various noise and error sources. This complexity increases in automotive contexts by vehicle movement and environmental factors influencing signal propagation, with multipath effects being particularly challenging to simulate accurately. This paper introduces a novel pipeline that leverages a 3D virtual environment to produce more realistic GNSS synthetic data for automotive applications. The pipeline integrates the CARLA Simulator and GPSoft’s SatNav Toolbox for Matlab, with custom-developed modules that generate raw GNSS measurements incorporating environment- and location-specific multipath effects. Our contributions include a tailored simulation pipeline for automotive applications, with integration of GNSS satellite orbits within CARLA, a dynamic multipath model reflecting obstacles in the simulated environment, and a synthetic dataset generated by this approach available to the community. Evaluation on CARLA’s Town03 map showed that while standard multipath models result in unrealistic uniform effects, our dynamic model produces effects that correlate with the vehicle’s surroundings, accurately reflecting real-world conditions such as increased errors in urban areas and lack of signals in tunnels. This approach can support the research, development, and validation of GNSS positioning algorithms and Artificial Intelligence (AI) model training, with potential applications extending also beyond the automotive context.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"13 ","pages":"35376-35386"},"PeriodicalIF":3.4,"publicationDate":"2025-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10893692","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143496542","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this paper, a new deployable grasping mechanism for non-cooperative space debris is proposed and developed. This grasping mechanism consists of three robotic fingers connected to a platform. Each finger is developed by combining a series of scissors mechanisms, in such a way that one mechanism drives the next. A half scissors mechanism is used at the end of finger as a its tip. These fingers are deployable and their length increases and decreases with the closing and opening of the scissors mechanism. Each deployable modules is equipped with a grasp driver mechanism, which can gradually bend the finger during the process of increase in its length, in order to accomplish the grasping of the non-cooperative space debris. Each finger is designed as an under-actuated mechanism, to save the development cost and make the finger lightweight. A special mechanism is developed in the platform of the grasping mechanism, such that single motor can be used to deploy and bend all the fingers, simultaneously. In the end, the validation of the working and effectiveness of the proposed deployable grasping mechanism is given through simulations and experimental work. It can be observed through the results that the proposed mechanism is able to grasp large objects with simultaneous deployment and bending of all fingers by using single motor.
{"title":"Design and Analysis of Single Motor-Driven Deployable Grasping Mechanism for Non-Cooperative Space Debris","authors":"Sajjad Manzoor;Yibo Wang;Kyungtae Kim;Qiang Lu;Youngjin Choi","doi":"10.1109/ACCESS.2025.3543731","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3543731","url":null,"abstract":"In this paper, a new deployable grasping mechanism for non-cooperative space debris is proposed and developed. This grasping mechanism consists of three robotic fingers connected to a platform. Each finger is developed by combining a series of scissors mechanisms, in such a way that one mechanism drives the next. A half scissors mechanism is used at the end of finger as a its tip. These fingers are deployable and their length increases and decreases with the closing and opening of the scissors mechanism. Each deployable modules is equipped with a grasp driver mechanism, which can gradually bend the finger during the process of increase in its length, in order to accomplish the grasping of the non-cooperative space debris. Each finger is designed as an under-actuated mechanism, to save the development cost and make the finger lightweight. A special mechanism is developed in the platform of the grasping mechanism, such that single motor can be used to deploy and bend all the fingers, simultaneously. In the end, the validation of the working and effectiveness of the proposed deployable grasping mechanism is given through simulations and experimental work. It can be observed through the results that the proposed mechanism is able to grasp large objects with simultaneous deployment and bending of all fingers by using single motor.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"13 ","pages":"33246-33258"},"PeriodicalIF":3.4,"publicationDate":"2025-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10892103","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143489156","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-19DOI: 10.1109/ACCESS.2025.3541480
Wenbo Xu;Rohana Mahmud;Wai Lam Hoo
Presents corrections to the paper, (Corrections to “A Systematic Literature Review: Are Automated Essay Scoring Systems Competent in Real-Life Education Scenarios?”).
{"title":"Corrections to “A Systematic Literature Review: Are Automated Essay Scoring Systems Competent in Real-Life Education Scenarios?”","authors":"Wenbo Xu;Rohana Mahmud;Wai Lam Hoo","doi":"10.1109/ACCESS.2025.3541480","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3541480","url":null,"abstract":"Presents corrections to the paper, (Corrections to “A Systematic Literature Review: Are Automated Essay Scoring Systems Competent in Real-Life Education Scenarios?”).","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"13 ","pages":"29738-29738"},"PeriodicalIF":3.4,"publicationDate":"2025-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10896632","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143471110","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-19DOI: 10.1109/ACCESS.2025.3543600
Romulo Gonçalves Lins;Tiago Nascimento de Freitas;Ricardo Gaspar
This paper presents a predictive maintenance (PdM) strategy for commercial vehicles, focusing on the turbocharger—a critical yet often under-monitored component. By combining sensor signals, workshop maintenance logs, and technical specifications, the study demonstrates how data-driven deep-learning techniques can robustly identify pending failures. Specifically, Long Short-Term Memory (LSTM) and Bidirectional LSTM (BiLSTM) architectures were employed to capture temporal dependencies and detect patterns that conventional approaches and purely onboard monitoring might overlook. Results on real-world fleet data indicate that BiLSTM achieved higher recall (98.65%) and a lower cost-score than standard LSTM, highlighting its effectiveness in minimizing missed failures. Although BiLSTM incurred slightly higher computational overhead, its superior performance underscores the value of integrating multi-sourced data and advanced sequence models for reliable, actionable PdM in heavy-duty fleets.
{"title":"Methodology for Commercial Vehicle Mechanical Systems Maintenance: Data-Driven and Deep-Learning-Based Prediction","authors":"Romulo Gonçalves Lins;Tiago Nascimento de Freitas;Ricardo Gaspar","doi":"10.1109/ACCESS.2025.3543600","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3543600","url":null,"abstract":"This paper presents a predictive maintenance (PdM) strategy for commercial vehicles, focusing on the turbocharger—a critical yet often under-monitored component. By combining sensor signals, workshop maintenance logs, and technical specifications, the study demonstrates how data-driven deep-learning techniques can robustly identify pending failures. Specifically, Long Short-Term Memory (LSTM) and Bidirectional LSTM (BiLSTM) architectures were employed to capture temporal dependencies and detect patterns that conventional approaches and purely onboard monitoring might overlook. Results on real-world fleet data indicate that BiLSTM achieved higher recall (98.65%) and a lower cost-score than standard LSTM, highlighting its effectiveness in minimizing missed failures. Although BiLSTM incurred slightly higher computational overhead, its superior performance underscores the value of integrating multi-sourced data and advanced sequence models for reliable, actionable PdM in heavy-duty fleets.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"13 ","pages":"33799-33812"},"PeriodicalIF":3.4,"publicationDate":"2025-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10892108","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143513006","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-19DOI: 10.1109/ACCESS.2025.3543749
Xiaoli Huan;Hong Zhou;Byungkwan Jung;Long Ma
Gait analysis is a critical tool for diagnosing and assessing the severity of Parkinson’s disease (PD). This study introduces a novel parallel hybrid architecture combining Conv1D, Efficient Transformers, and Bidirectional GRU layers to analyze gait data for both PD detection and severity staging. Conv1D layers extract local spatial features, Efficient Transformers capture contextual dependencies, and Bidirectional GRUs model temporal patterns in VGRF signals. Designed to balance computational efficiency and scalability, the model demonstrates state-of-the-art performance, achieving 95.7% accuracy in PD detection and 87.3% accuracy in severity staging on the PhysioNet gait dataset. Additionally, the architecture is highly versatile, offering potential for application in other 1D signal analysis tasks.
{"title":"Enhancing Gait Analysis for Parkinson’s Disease Detection and Severity Staging With a Parallel Conv1D-Efficient Transformer and Bidirectional GRU Hybrid Architecture","authors":"Xiaoli Huan;Hong Zhou;Byungkwan Jung;Long Ma","doi":"10.1109/ACCESS.2025.3543749","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3543749","url":null,"abstract":"Gait analysis is a critical tool for diagnosing and assessing the severity of Parkinson’s disease (PD). This study introduces a novel parallel hybrid architecture combining Conv1D, Efficient Transformers, and Bidirectional GRU layers to analyze gait data for both PD detection and severity staging. Conv1D layers extract local spatial features, Efficient Transformers capture contextual dependencies, and Bidirectional GRUs model temporal patterns in VGRF signals. Designed to balance computational efficiency and scalability, the model demonstrates state-of-the-art performance, achieving 95.7% accuracy in PD detection and 87.3% accuracy in severity staging on the PhysioNet gait dataset. Additionally, the architecture is highly versatile, offering potential for application in other 1D signal analysis tasks.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"13 ","pages":"33351-33360"},"PeriodicalIF":3.4,"publicationDate":"2025-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10892107","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143489199","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}