Computers & Electrical Engineering最新文献

{"title":"High-performance and non-contact energy harvesters from high-voltage power lines magnetic fields","authors":"Bahram Rashidi","doi":"10.1016/j.compeleceng.2025.110267","DOIUrl":"10.1016/j.compeleceng.2025.110267","url":null,"abstract":"<div><div>In this paper, high-performance and non-contact energy harvesters from the magnetic fields of high-voltage power lines are presented. The energy harvesters are based on a rod ferrite core in the middle of the coil and two <span><math><mi>⊐</mi></math></span>-shaped ferrite cores at both ends of the rod ferrite core. The use of two <span><math><mi>⊐</mi></math></span>-shaped ferrite cores can provide several times improvement in power. Exposing these cores to magnetic flux increases the magnetic flux guiding efficiency and increases the energy harvesting rate. Here, by connecting a voltage rectifier circuit, we design a portable power supply based on energy harvesting from magnetic fields, which can provide the power consumption needs of some low-power circuits. The results of the capacitor charging ability, the open-circuit voltage across the harvesters, and the output power are measured to evaluate the performance of the proposed structures. The results show that the structures exhibit acceptable performance for practical requirements. Based on the results, a maximum open-circuit voltage of 13.619 V and an output power of 4.449 mW under a magnetic field of 7 <span><math><mi>μ</mi></math></span>T are achieved for the best proposed structure. Considering that the proposed energy harvesting structures are non-contact and low-cost, therefore, they can be used to provide power for some low-power wireless monitoring sensors in high-voltage power transmission systems.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110267"},"PeriodicalIF":4.0,"publicationDate":"2025-03-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143679517","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":"TraceAwareness and dual-strategy fuzz testing: Enhancing path coverage and crash localization with stochastic science and large language models","authors":"Xiaoquan Chen ,&nbsp;Jian Liu ,&nbsp;Yingkai Zhang ,&nbsp;Qinsong Hu ,&nbsp;Yupeng Han ,&nbsp;Ruqi Zhang ,&nbsp;Jingqi Ran ,&nbsp;Lei Yan ,&nbsp;Baiqi Huang ,&nbsp;Shengtin Ma","doi":"10.1016/j.compeleceng.2025.110266","DOIUrl":"10.1016/j.compeleceng.2025.110266","url":null,"abstract":"<div><div>This paper proposes an innovative fuzzing technique to address path coverage and crash localization challenges inherent in traditional methods. We introduce TraceAwareness, a technology for precise tracking and recording of program execution paths, significantly enhancing fuzzing efficiency and issue traceability. Additionally, we present a dual-strategy method (DSM-SST-LLMT) based on stochastic science theory and large language model technology, combining random exploration with intelligent analysis for effective test input generation. Experimental evaluations demonstrate that our technique achieves 85% edge coverage compared to AFL++’s 35%, discovers 3,000 new paths versus AFL++’s 800, and identifies 8 critical crashes where AFL++ found none. Our approach shows particular strength in handling complex and diverse inputs, reaching 2-3 times the maximum path depth of AFL++. This research offers new directions for improving software testing efficiency and reliability, with potential applications in critical infrastructure, cloud-based systems, and IoT environments.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110266"},"PeriodicalIF":4.0,"publicationDate":"2025-03-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143679513","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":"Embedded deep learning models for multilingual speech recognition","authors":"Mohamed Hedi Rahmouni ,&nbsp;Mohamed Salah Salhi ,&nbsp;Ezzeddine Touti ,&nbsp;Hatem Allagui ,&nbsp;Mouloud Aoudia ,&nbsp;Mohammad Barr","doi":"10.1016/j.compeleceng.2025.110271","DOIUrl":"10.1016/j.compeleceng.2025.110271","url":null,"abstract":"<div><div>This paper investigates the hybridization of Genetic Algorithms (GA) with Recurrent Self-Organizing Maps (RSOM) for speech recognition. It ensures the benchmarking of its performance against traditional and deep learning-based methods, including Hidden Markov Models (HMM), Support Vector Machines (SVM), Convolutional Neural Networks (CNN), Recurrent Neural Networks (RNN), Long Short-Term Memory networks (LSTM), Gated Recurrent Units (GRU), and wave to vector 2.0 (wav2vec 2.0). The aim of this study is to demonstrate the performance of the hybrid GA-RSOM model implemented on an embedded system, such as a modern Digital Signal Processing (DSP). The evaluation is carried out in terms of reaction time and recognition accuracy for speech with very high variability and multilingual content. Experiments show that while the GARSOM model is slower than some models like CNN, it achieves a stable and precise recognition rate of up to 98 %, depending on the phonemes.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110271"},"PeriodicalIF":4.0,"publicationDate":"2025-03-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143679509","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 coupled framework for power load forecasting with Gaussian implicit spatio temporal block and attention mechanisms network","authors":"Dezhi Liu ,&nbsp;Xuan Lin ,&nbsp;Hanyang Liu ,&nbsp;Jiaming Zhu ,&nbsp;Huayou Chen","doi":"10.1016/j.compeleceng.2025.110263","DOIUrl":"10.1016/j.compeleceng.2025.110263","url":null,"abstract":"<div><div>The increasing demand for electricity underscores the need for accurate power load forecasting to optimize grid management and resource allocation. With the emergence of more complex multi-energy hybrid systems, the resulting multivariate power load data pose significant challenges for precise forecasting. To address this, we propose a novel framework that integrates Variational Mode Decomposition (VMD) with an Encoder–Decoder architecture featuring customized Gaussian Implicit Spatio-Temporal (GIST) blocks to uncover implicit spatial dependencies across temporal and multi-feature dimensions. Initially, VMD decomposes the original time series into multiple resolution components, effectively reducing noise and extracting intrinsic temporal patterns. These components are then processed by an Encoder–Decoder network for prediction. Within each GIST block, token embedding is applied to the input before being fed into a Gaussian Mixture Model (GMM)-based implicit spatio-temporal representation module. Unlike conventional expectation–maximization (EM) algorithms, our learned Gaussian modeling approach provides a more adaptive and computationally efficient alternative for residential power load forecasting. Temporal dependencies are further captured through Long Short-Term Memory (LSTM) units and attention mechanisms across subsequent blocks, enhancing the model’s predictive capability. Experimental validation demonstrates the superior performance of our proposed model, achieving reductions of 7.98% and 9.32% in Mean Absolute Error (MAE) and Root Mean Square Error (RMSE), respectively, compared to existing forecasting models. Notably, our GMM-based approach outperforms traditional two-dimensional convolution-based methods, yielding improvements of 11.3% and 5.72% in MAE and RMSE, highlighting the efficacy of our framework in handling complex multivariate power load data.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110263"},"PeriodicalIF":4.0,"publicationDate":"2025-03-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143679511","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":"Revolutionizing power electronics design through large language models: Applications and future directions","authors":"Khalifa Aliyu Ibrahim ,&nbsp;Patrick Chi-Kwong Luk ,&nbsp;Zhenhua Luo ,&nbsp;Seng Yim Ng ,&nbsp;Lee Harrison","doi":"10.1016/j.compeleceng.2025.110248","DOIUrl":"10.1016/j.compeleceng.2025.110248","url":null,"abstract":"<div><div>The design of electronic circuits is critical for a wide range of applications, from the electrification of transportation to the Internet of Things (IoT). It demands substantial resources, is time-intensive, and can be highly intricate. Current design methods often lead to inefficiencies, prolonged design cycles, and susceptibility to human error. Advancements in artificial intelligence (AI) play a crucial role in power electronics design by increasing efficiency, promoting automation, and enhancing sustainability of electrical systems. Research has demonstrated the applications of AI in power electronics to enhance system performance, optimization, and control strategy using machine learning, fuzzy logic, expert systems, and metaheuristic methods. However, a review that includes the recent AI advancements and potential of large language models (LLMs) like generative pre-train transformers (GPT) has not been reported. This paper presents an overview of applications of AI in power electronics (PE) including the potential of LLMs. The influence of LLMs-AI on the design process of PE and future research directions is also highlighted. The development of advanced AI algorithms such as pre-train transformers, real-time implementations, interdisciplinary collaboration, and data-driven approaches are also discussed. The proposed LLMs-AI is used to design parameters of high-frequency wireless power transfer (HFWPT) using MATLAB as a first case study, and high-frequency alternating current (HFAC) inverter using PSIM as a second case study. The proposed LLM-AI driven design is verified based on a similar design reported in the literature and Wilcoxon signed-rank test was conducted to further validate the result. Results show that the LLM-AI driven design based on the OpenAI foundation model has the potential to streamline the design process of power electronics. These findings provide a good reference on the feasibility of LLMs-AI on power electronic design.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110248"},"PeriodicalIF":4.0,"publicationDate":"2025-03-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143679510","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}

{"title":"Towards a physical imaging-driven sparse attention dehazer for Internet of Things-aided Maritime Intelligent Transportation","authors":"Yuxuan Tian ,&nbsp;Yu Guo ,&nbsp;Yuxu Lu ,&nbsp;Yuan Gao ,&nbsp;Ryan Wen Liu","doi":"10.1016/j.compeleceng.2025.110257","DOIUrl":"10.1016/j.compeleceng.2025.110257","url":null,"abstract":"<div><div>In the field of Maritime Intelligent Transportation Systems (MITS), the integration of Internet of Things (IoT) technologies and intelligent algorithms has revolutionized visual IoT-aided MITS. This integration, enabled by advanced communication technologies, network infrastructures, sensor capabilities, and data science methodologies, has significantly enhanced monitoring, navigation, and collision avoidance systems, thus improving waterway transportation efficiency. However, the performance of these systems can be hampered by atmospheric conditions, leading to degraded imaging quality characterized by contrast reduction, color distortion, and object invisibility. Such challenges impede critical vision-based tasks like object detection, tracking, and scene understanding in MITS. To address the performance gap between clear and hazy scenes, we propose a novel framework called PSDformer. This framework integrates Top-K Sparse Attention with a Physics-Aware Feed-Forward Network to enhance performance under hazy conditions. Additionally, we introduce a novel paired data generation method to reduce the disparity between synthetic and real-world data. Experimental results on synthetic and real-world datasets demonstrate that PSDformer outperforms existing state-of-the-art methods in both qualitative and quantitative evaluations. Importantly, its exceptional dehazing capability significantly improves detection accuracy under adverse hazy conditions, thereby addressing a critical challenge in visual IoT-aided MITS.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110257"},"PeriodicalIF":4.0,"publicationDate":"2025-03-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143679518","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":"Cross spatial and Cross-Scale Swin Transformer for fine-grained age estimation","authors":"Linbu Xu,&nbsp;Chunlong Hu,&nbsp;Xin Shu,&nbsp;Hualong Yu","doi":"10.1016/j.compeleceng.2025.110264","DOIUrl":"10.1016/j.compeleceng.2025.110264","url":null,"abstract":"<div><div>Facial age estimation is a classic problem in the field of computer vision. Previous studies have shown that learning discriminative features is crucial for accurate age estimation. Although Swin Transformer has been successfully applied on many computer vision tasks, it cannot effectively capture directional features during the aging process for age estimation task. Moreover, it still exhibits bias towards global features and cannot capture more fine-grained age-related features, ultimately leading to ambiguity in distinguishing adjacent ages. To address these issues, we propose <strong>Cross Spatial and Cross-Scale Swin Transformer (CSCS-Swin)</strong> that can extract fine-grained age-related features. Firstly, the <strong>Cross Spatial Feature Block (CSFB)</strong> module is constructed in CSCS-Swin, which extracts facial wrinkle features and craniofacial features along the horizontal and vertical directions, and models feature associations between different facial regions. Secondly, considering that the discrimination power of features at different scales differs in facial regions, <strong>Cross-Scale Feature Partition (CSFP)</strong> is proposed, which can precisely extract corss-scale fine-grained features. Lastly, the <strong>Feature Enhancement Module (FEM)</strong> is introduced to further enhance the ability of feature representation. These three modules in CSCS-Swin work together to improve the accuracy of age estimation. Extensive experiments on four popular datasets, namely, MORPH II, UTKFace, AFAD, and CACD, demonstrate the superiority of the proposed method.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110264"},"PeriodicalIF":4.0,"publicationDate":"2025-03-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143679512","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":"Time–Frequency Domain Joint Noise Reduction Multi-Resolution Power System Time Series Prediction Network","authors":"Ziyang Cao ,&nbsp;Hangwei Tian ,&nbsp;Qihao Xu ,&nbsp;Jinzhu Wang","doi":"10.1016/j.compeleceng.2025.110255","DOIUrl":"10.1016/j.compeleceng.2025.110255","url":null,"abstract":"<div><div>This paper proposes an innovative time series prediction method designed for power systems to overcome the shortcomings of existing deep learning techniques in complex noise environments. The method, called <strong>T</strong>ime–Frequency <strong>D</strong>omain Joint Noise <strong>R</strong>eduction Multi-Resolution Power <strong>S</strong>ystem Time S<strong>e</strong>ries Prediction Network (TDRSE), investigates the impact of noise on prediction results and proposes a complete solution. TDRSE consists of two key components Exponential Decay-based Denoising Network (EDnet) and Dynamic Frequency-Domain Signal Enhancement Network (FDse). EDnet achieves dynamic attention to different time points in the time dimension by introducing exponential decay units to cope with the volatility of power loads and noise disturbances. At the same time, FDse employs frequency-domain enhancement techniques and adaptive thresholding strategies to remove the noise components in the frequency domain, thus further improving the model’s power data prediction accuracy. The experimental results show that TDRSE performs well on real data sets of multiple power systems, significantly improves the prediction accuracy under complex noise conditions, and reaches the industry-leading level.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110255"},"PeriodicalIF":4.0,"publicationDate":"2025-03-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143679040","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":"Single stage transformer less multilevel inverter for solar PV application","authors":"Vinayak Kumar,&nbsp;Ruchi Agarwal","doi":"10.1016/j.compeleceng.2025.110243","DOIUrl":"10.1016/j.compeleceng.2025.110243","url":null,"abstract":"<div><div>The article presents a single stage transformer less multilevel inverter (SSTL-MLI) with common ground based inverter topology for grid tied PV application. It is designed with only 5 switches for 5 level generation. It uses switched capacitor approach therefore, needs single DC-source. Moreover, the topology has self voltage balancing feature, thereby no need of additional circuit. The peak current control (PCC) strategy is utilized to regulate switching pulse. MATLAB/Simulink is used to assess the proposed topology performance. It results voltage and current total harmonic distortion (THD) with values 32.49% and 1.66% respectively at the specified load at inverter output. It also satisfies IEEE 519 standard for power quality with grid current THD with value 2.12% for 20 harmonic order. The proposed system obtains maximum efficiency (<span><math><mi>η</mi></math></span>) with 98.30% value at input voltage (400V), and output power (4.72 kW). The cost function, composition with number of switch counts, DC source, weight coefficient and total standing voltage value has lowest value i.e. 16.75 in the proposed topology. Moreover, a scaled down prototype with 350 W rating is built of the proposed topology and validated under number of test condition such as change of load i.e resistive to inductive load, change of modulation index, and frequency change condition. The detailed comparative analysis is also presented in terms of switch count, TVS, MSV, efficiency, gain, cost function by comparing of number of existing topology so that the favorable feature of the proposed topology can be highlighted.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110243"},"PeriodicalIF":4.0,"publicationDate":"2025-03-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143679514","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":"Harnessing hybrid intelligence: Four vector metaheuristic and differential evolution for optimized photovoltaic parameter extraction","authors":"Charaf Chermite,&nbsp;Moulay Rachid Douiri","doi":"10.1016/j.compeleceng.2025.110276","DOIUrl":"10.1016/j.compeleceng.2025.110276","url":null,"abstract":"<div><div>Accurate parameter extraction in photovoltaic (PV) cells and modules is crucial for optimizing performance and ensuring efficient energy conversion in solar technologies. However, existing optimization methods exhibit inherent limitations. The Four Vector Intelligent Metaheuristic (FVIM) demonstrates strong local refinement but suffers from limited global exploration and premature convergence. Meanwhile, Differential Evolution (DE) offers effective global search but often struggles with stagnation in local optima. To overcome these challenges, we introduce a novel hybrid algorithm that synergistically combines FVIM's multi-vector refinement strategy with DE's robust mutation and crossover mechanisms. This hybridization ensures a balanced trade-off between local exploitation and global exploration, significantly reducing the Root Mean Square Error (RMSE) between measured and estimated current values, ensuring precise parameter estimation. The FVIM-DE algorithm is rigorously benchmarked against 15 state-of-the-art metaheuristic algorithms across three standard PV models: the Single Diode Model (SDM), Double Diode Model (DDM), and Photovoltaic Module Model (PMM). Additionally, it was evaluated on various PV technologies under different irradiance and temperature conditions. Additionally, it was evaluated on various PV technologies under different irradiance and temperature conditions. FVIM-DE consistently achieves the lowest RMSE values, with a minimum of 9.8602E-4 for SDM, 9.8248E-4 for DDM, and 2.4250E-3 for PMM, surpassing all competing algorithms. Furthermore, the Friedman test ranks FVIM-DE first across all PV models, highlighting its robustness and statistical superiority. Results consistently highlight FVIM-DE's superior accuracy, rapid convergence, and adaptability, outperforming other methods in minimizing RMSE. This positions FVIM-DE as a reliable and effective tool for PV parameter extraction, advancing solar energy applications under diverse environmental conditions.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110276"},"PeriodicalIF":4.0,"publicationDate":"2025-03-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143679508","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}