Computers & Electrical Engineering最新文献

{"title":"A 17-level octuple boost inverter with low component and reduced inrush current","authors":"Hossein Mansourizadeh ,&nbsp;Majid Hosseinpour ,&nbsp;Ali Seifi ,&nbsp;Hamed Heydari-doostabad","doi":"10.1016/j.compeleceng.2025.110586","DOIUrl":"10.1016/j.compeleceng.2025.110586","url":null,"abstract":"<div><div>This paper proposes a 17-level switched-capacitor-based inverter structure to reduce circuit components, limit inrush current, and provide an eightfold voltage-boosting capability. The proposed 17-level inverter comprises eleven switches, four capacitors, and three diodes. The distinct advantages of the proposed structure include the requirement for fewer components, inrush current limiting capability, eight-times voltage boosting, and a reduced number of components with a maximum blocking voltage (MBV) applied to only two switches. The circuit of the proposed structure includes a series-parallel unit (SPSC), two level-doubler (LD) units, and two half-bridge (HB) units for voltage polarity reversal. The circuit operation, charging and discharging states, the proposed modulation method, and the charging current analysis are presented. The capacitor voltages in the proposed structure are automatically balanced without needing auxiliary or pre-charging circuits. The proposed inverter is compared with other topologies from recent studies based on various parameters, demonstrating the superiority of the proposed design. The proposed 17-level structure simulation has been examined under steady-state and dynamic conditions. Ultimately, the experimental findings validate the theoretical assessments and simulation results.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110586"},"PeriodicalIF":4.0,"publicationDate":"2025-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144670927","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}

{"title":"Hybrid deep learning system for crop disease classification using modified SegNet segmentation","authors":"Mukesh Kumar Tripathi ,&nbsp;D.N. Vasundhara ,&nbsp;V.K.N.S.N. Moorthy Ch ,&nbsp;Kapil Misal ,&nbsp;Bhagyashree Ashok Tingare ,&nbsp;Sanjeevkumar Angadi","doi":"10.1016/j.compeleceng.2025.110576","DOIUrl":"10.1016/j.compeleceng.2025.110576","url":null,"abstract":"<div><div>In traditional agricultural systems, managing crop diseases faces significant challenges, primarily due to the reliance on visual inspection and manual symptom identification. These methods are often time-consuming, error-prone, and may fail to detect diseases early or accurately, leading to ineffective treatments and substantial crop loss. Furthermore, the unpredictability of disease symptoms and the existence of similar-looking diseases complicate diagnosis. To address these limitations, there is a growing necessity for innovative deep learning-based methods. This study proposes an advanced Modified LinkNet-Bidirectional Long Short-Term Memory (MLBLSTM)-based system for crop disease classification, incorporating a multi-step process starting with data collection from three datasets: apple, corn, and pepper plant leaves. The preprocessing phase utilizes Enhanced Wiener Filtering (EWF) to preserve high-frequency details and enhance image quality. The filtered images are processed through an advanced Modified SegNet (MSegNet) model to do the segmentation process. Feature extraction follows, leveraging Hierarchy of Skeleton (HOS), Modified Local Gabor Increasing Pattern (MLGIP), Median Binary Pattern (MBP), and statistical features. Finally, the classification step employs a hybrid model combining Modified LinkNet (MLNet) with a novel σ-SE block and Bidirectional Long Short-Term Memory (Bi-LSTM) classifiers. The validation results prove the performance of MLBLSTM model measures with an accuracy of 0.947, a sensitivity of 0.955, and a specificity of 0.936.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110576"},"PeriodicalIF":4.0,"publicationDate":"2025-07-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144665658","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}

{"title":"3DMalDroid: A novel 3D image based approach for android malware detection and classification","authors":"Muhammed Mutlu Yapici","doi":"10.1016/j.compeleceng.2025.110542","DOIUrl":"10.1016/j.compeleceng.2025.110542","url":null,"abstract":"<div><div>Android is one of the most widely preferred and utilized operating systems today. Consequently, it has attracted the attention of hackers, and Android device users are increasingly subjected to cyberattacks. This study aims to develop a solution for malware attacks targeting Android-based devices. To achieve this, we propose two novel deep learning-based systems that utilize 2D+ and 3D images for malware detection and malware category classification. The system yielding the best results, which is based on 3D imaging, is named 3DMalDroid. Furthermore, we address imbalanced data and duplicated data issues, which contribute to bias and overfitting in malware detection and classification results. The results demonstrate that the proposed 3DMalDroid system surpasses state-of-the-art studies in the literature, achieving an accuracy of 0.994, precision of 0.993, recall of 0.992, and an F1-score of 0.993. In conclusion, the proposed 3DMalDroid system makes a significant contribution to Android malware detection by addressing duplicate data and class imbalance issues.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110542"},"PeriodicalIF":4.0,"publicationDate":"2025-07-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144653504","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}

{"title":"Feature fusion based automatic chord recognition model: BTC-FDAA-FGF","authors":"Chen Li,&nbsp;Hao Wu,&nbsp;JingYi Jiang,&nbsp;Lihua Tian","doi":"10.1016/j.compeleceng.2025.110555","DOIUrl":"10.1016/j.compeleceng.2025.110555","url":null,"abstract":"<div><div>Automatic chord recognition is a significant topic in the field of Music Information Retrieval (MIR). This paper introduces a novel feature fusion method combining Hybrid Constant-Q Transform (HCQT) and adaptive attention for chord detection, especially with a focus on improving the accuracy of chord detection of rare chord classes. Serving as one of the cornerstone features of music, the chords obtained by chord recognition algorithms are the basis of many high-level semantic tasks. At present, a severe class imbalance problem exists in the domain of automatic chord recognition. The recognition accuracy of rare chords is much lower than that of common chords, which significantly affect the overall performance of chord recognition algorithms. In this paper, a chord recognition algorithm based on feature fusion is proposed. First, in the feature extraction part, Hybrid Constant-Q Transform (HCQT) is introduced to assist with Constant-Q Transform(CQT) to obtain richer and finer musical signal features, enabling better tracking of overtones. Next, in the chord estimation part, the frequency-domain adaptive attention (FDAA) mechanism is used to enhance feature saliency, ensuring that the network can adaptively adjust the weights for different frequency components when training. Thereby frequency-domain features that contain important information can be selectively enhanced. The enhanced features are then fed into an aggregation module that integrates a bidirectional self-attention module and Fourier transform module, enabling more effective capture of fine-grained features, global context information, and periodic structures in chords. The experimental result shows that proposed algorithm outperforms existing mainstream baseline methods by 1.2% to 2.2% on the MIREX metrics, validating the effectiveness of the algorithm.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110555"},"PeriodicalIF":4.0,"publicationDate":"2025-07-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144634034","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}

{"title":"A lightweight intrusion detection system using deep convolutional neural network","authors":"Vanlalruata Hnamte ,&nbsp;Ashfaq Ahmad Najar ,&nbsp;Chhakchhuak Laldinsanga ,&nbsp;Jamal Hussain ,&nbsp;Lal Hmingliana","doi":"10.1016/j.compeleceng.2025.110561","DOIUrl":"10.1016/j.compeleceng.2025.110561","url":null,"abstract":"<div><div>Intrusion Detection Systems (IDSs) serve as critical components of cybersecurity infrastructure, safeguarding computer networks against evolving cyber threats. Recent advancements in deep learning architectures, particularly Convolutional Neural Networks (CNNs), have demonstrated substantial potential in augmenting IDS efficacy. However, conventional CNN architectures exhibit inherent limitations in processing sequential data due to their inability to capture long-term temporal dependencies. To address these operational constraints, this study proposes a lightweight deep convolutional neural network-based intrusion detection system (LWIDS-DCNN), a novel framework designed to optimize feature extraction and detection accuracy in heterogeneous network environments. The LWIDS-DCNN architecture strategically integrates convolutional layers with pooling operations and fully connected layers, forming an optimized algorithmic structure tailored for efficient extraction of discriminative features from network traffic data. The framework incorporates adaptive accelerator algorithms and dynamic learning rate optimization strategies to ensure accelerated convergence rates while maintaining training stability. Empirical validation was conducted using three benchmark datasets: CICIDS2017, CICIoMT2024, and InSDN. The proposed model achieved state-of-the-art detection accuracy, with results exceeding 99.93% on CICIDS2017, 99.70% on CICIoMT2024, and 99.97% on InSDN. A comprehensive comparative analysis against existing methodologies demonstrated LWIDS-DCNN’s superiority across key performance metrics, including precision, recall, F1-score, and loss rate. Notably, the system’s lightweight design ensures computational efficiency without compromising detection robustness, making it particularly suitable for resource-constrained environments. This work contributes to the advancement of network security research by introducing a scalable, high-performance IDS architecture capable of addressing the unique challenges posed by traditional networks, IoMT ecosystems, and SDN infrastructures. The LWIDS-DCNN framework establishes a foundational paradigm for real-time intrusion detection in converged environments, offering a robust, lightweight solution that addresses the unique intrusion detection requirements of emerging IoT and SDN systems.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110561"},"PeriodicalIF":4.0,"publicationDate":"2025-07-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144633994","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}

{"title":"Remote heart rate measurement based on spatial–temporal self-attention","authors":"Ning Sun ,&nbsp;Ningbin Wang ,&nbsp;Jixin Liu ,&nbsp;Lei Chai ,&nbsp;Haian Sun","doi":"10.1016/j.compeleceng.2025.110557","DOIUrl":"10.1016/j.compeleceng.2025.110557","url":null,"abstract":"<div><div>Remote photoplethysmography (rPPG), a non-contact technique for measuring heart rate, has gained significant traction due to its convenience and non-invasive nature. However, rPPG signals are inherently weak and exhibit regional variations across facial areas. Accurate heart rate estimation necessitates analysis of extended-duration video sequences (exceeding 100 frames). To address these challenges, this paper proposes a novel deep neural network (DNN) based on self-attention mechanism, named spatial–temporal self-attention network (STSA-Net). This convolution-free DNN primarily adopts a transformer architecture. Initially, differential processing is applied to the input video to accentuate frame-to-frame variations and amplify subtle rPPG signals. These differential frames are subsequently passed through a spatial self-attention encoding module. This module models spatial dependencies within facial regions, allowing the network to focus on informative areas while suppressing irrelevant noise. Following spatial encoding, the features are processed by a temporal self-attention module, which captures long-term dependencies across frames using transformer-based techniques. The proposed method is rigorously evaluated through comprehensive experiments, including ablation studies, intra-database evaluations, and cross-database comparisons, using three benchmark databases: UBFC-RPPG, PURE, and MAHNOB-HCI. Our model demonstrates performance on par with state-of-the-art methods for remote heart rate measurement, achieving a MAE of 0.4 bpm on PURE, 0.48 bpm on UBFC, and 3.75 bpm on MAHNOB-HCI in intra-database experiments, and 1.36 bpm (UBFC<span><math><mo>→</mo></math></span> PURE) and 1.27 bpm (PURE<span><math><mo>→</mo></math></span> UBFC) in cross-database experiments. Additionally, an outlier in the PURE database is identified, and its cause and impact are analyzed.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110557"},"PeriodicalIF":4.0,"publicationDate":"2025-07-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144633993","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}

{"title":"A power quality disturbance classification method based on multi-level optimized adaptive time-frequency analysis and deep learning","authors":"Shaoyang Wang ,&nbsp;Qiang Zhang ,&nbsp;Menghan Li ,&nbsp;Zekai Zhao ,&nbsp;Ce Cao","doi":"10.1016/j.compeleceng.2025.110548","DOIUrl":"10.1016/j.compeleceng.2025.110548","url":null,"abstract":"<div><div>With the increasing integration of distributed generation in the power grid and the growing power quality challenges caused by various high-frequency power electronic devices, the stability of the grid is increasingly affected. In this paper, a power quality disturbance (PQD) recognition method based on a combination of time-frequency analysis and Deep Convolutional Neural Networks (DCNN) is proposed. First, the Variational Mode Decomposition (VMD) parameters are optimized using the Mean Filter Envelope Extremum Method (FE) and the Sparrow Search Algorithm (SSA). This is followed by the screening and removal of noise components from the modal components obtained through VMD decomposition. Secondly, for the selected multi-scale modal components, the Multi-Resolution S-Transform (MST) is applied to obtain the time-frequency feature maps of multiple scales, which are then input into an improved Inception-ResNet model for PQD recognition. Additionally, the Convolutional Block Attention Module (CBAM) attention mechanism is introduced to improve the model performance in capturing key information. The proposed method obtained a classification accuracy of 99.31 % in a 20–40 dB random noise environments and is validated through measured PQD data, demonstrating its reliability.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110548"},"PeriodicalIF":4.0,"publicationDate":"2025-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144632220","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}

{"title":"From classical to intelligent control: Evolving trends in robotic manipulator technology","authors":"Saba Waseem ,&nbsp;Muhammad Adnan ,&nbsp;Muhammad Sajid Iqbal ,&nbsp;Arslan Ahmed Amin ,&nbsp;Anwar Shah ,&nbsp;Muhammad Tariq","doi":"10.1016/j.compeleceng.2025.110559","DOIUrl":"10.1016/j.compeleceng.2025.110559","url":null,"abstract":"<div><div>Numerous industrial, medical, and scientific applications currently depend on robotic manipulators, which are essential to the expanding robotics sector. These manipulators cannot operate well without advanced control algorithms that consider precision, adaptability, and flexibility. This work is focused on listing many types, applications, and functions of control strategies for robotic manipulators. The review commences with a broad overview of manipulators, including subjects such as cable-driven, parallel, and serial models. It subsequently contextualizes the unique mechanical and functional characteristics of each type. The next sections encompass a range of control procedures, from conventional techniques such as Proportional-Integral-Derivative (PID) and linear control to complex systems that integrate hybrid control frameworks like machine learning (ML), and artificial intelligence (AI). This comprehensive analysis addresses some enduring issues like task flexibility, computational efficiency, and environmental uncertainty, together with the advantages and disadvantages of each approach. This study also emphasizes emergent domains, including federated learning (FL), blockchain integration, and quantum computing, while also identifying prospective future research topics. By combining current information and outlining potential developments, the study provides researchers and engineers with a reference for improving the control of robotic manipulators, hence improving performance and reliability in challenging working environments. The findings of this work demonstrate the efficacy of advanced control methodologies and pave the way for advances that could transform robotic manipulation in the future.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110559"},"PeriodicalIF":4.0,"publicationDate":"2025-07-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144614101","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}

{"title":"Integration of VCNN models with RVFLwoDL to boost the parking space classification","authors":"Navpreet,&nbsp;Rajendra Kumar Roul,&nbsp;Rinkle Rani","doi":"10.1016/j.compeleceng.2025.110444","DOIUrl":"10.1016/j.compeleceng.2025.110444","url":null,"abstract":"<div><div>Parking space classification is critical in elevating traffic congestion, reducing air pol- lution, and enhancing drivers’ convenience. This study introduces a robust model for parking space classification, ingeniously combining variants of the Convolutional Neu- ral Network (VCNN) with the Incremental Random Vector Function Link without di- rect link (I-RVFLwoDL). The primary innovation consists of substituting the fully connected layer of the VCNN with I-RVFLwoDL. This eliminates the need for a costly backpropagation procedure, resulting in a substantial decrease in training time. The integration of VCNN with I-RVFLwoDL utilizes the I-RVFLwoDL’s rapid learning efficency and robust generalization capabilities. I-RVFLwoDL simplifies the network structure by eliminating the complex neuron pathways typically found in other established methodologies. The proposed hybrid model’s effectiveness is rigorously as- sessed utilizing three established datasets: PKLot, CNRPark, and CNRPark+EXT. The system’s ability to distinguish between occupied and vacant parking spaces is demon- strated through various performance metrics used in machine learning. The proposed model’s performance is also evaluated against existing deep learning models to illustrate its superiority. This research presents significant potential for intelligent transportation systems, providing an efficient solution for parking space classification.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110444"},"PeriodicalIF":4.0,"publicationDate":"2025-07-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144614102","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}

{"title":"Image encryption scheme based on 2D-ICCM and bit-planes cross permutation-diffusion using parallel computing","authors":"Xingbin Liu,&nbsp;Shuyi Zheng,&nbsp;Jing Yang","doi":"10.1016/j.compeleceng.2025.110569","DOIUrl":"10.1016/j.compeleceng.2025.110569","url":null,"abstract":"<div><div>With the increasing reliance on digital images for communication and storage, image security has become a critical concern. Traditional encryption schemes often face challenges regarding efficiency and robustness against modern cryptographic attacks. In this paper, a novel image encryption scheme is proposed based on an innovative chaotic system named 2D-ICCM and cross bit-plane permutation, enhanced by parallel computing for improved efficiency. The 2D-ICCM is designed to exhibit complex dynamic properties and high sensitivity to initial conditions, the chaotic behaviors of which are verified through the Lyapunov exponent, bifurcation diagrams, sample entropy, and 0-1 test. In addition, a cross bit-plane permutation method is proposed to rearrange the bits in high four bit-planes and low four bit-planes, which further increases the complexity of the encryption and provides stronger protection against attacks. In the diffusion process, an independent bit-plane multi-direction diffusion method using a Zigzag scan is proposed. To address the issue of encryption speed, the parallel computing technique is applied for the diffusion process. Experimental results show that the proposed scheme offers high encryption quality, robust resistance to statistical analysis attacks, and significantly improved encryption speed compared to traditional methods.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110569"},"PeriodicalIF":4.0,"publicationDate":"2025-07-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144614100","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}