Computers & Electrical Engineering最新文献

{"title":"A federated learning model with the whale optimization algorithm for renewable energy prediction","authors":"Viorica Rozina Chifu,&nbsp;Tudor Cioara,&nbsp;Cristian Daniel Anitei,&nbsp;Cristina Bianca Pop,&nbsp;Ionut Anghel,&nbsp;Liana Toderean","doi":"10.1016/j.compeleceng.2025.110259","DOIUrl":"10.1016/j.compeleceng.2025.110259","url":null,"abstract":"<div><div>Federated prediction models for energy prosumers create a global model by combining insights from local machine learning models trained on-site without centralizing the data. For time series energy data, this collaborative approach faces challenges due to the non-IID nature of the data, variations in generation patterns, the high number of model parameters, and convergence issues, leading to poor prediction accuracy. This paper introduces a novel federated learning model, FedWOA, which uses the whale optimization algorithm to determine optimal aggregation coefficients based on the local model weight vectors by pondering the updates considering the model performance and data dimensionality construct the global shared model. To handle the non-IID data the prosumers were clustered based on the similarity of their energy profiles using K-Means. FedWOA improves the prediction quality at the prosumer site, with a 16 % average reduction of the mean absolute error compared to FedAVG while demonstrating good convergence and reduced loss.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110259"},"PeriodicalIF":4.0,"publicationDate":"2025-03-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143644962","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":"Privacy-Preserving Federated Learning with Differentially Private Hyperdimensional Computing","authors":"Fardin Jalil Piran ,&nbsp;Zhiling Chen ,&nbsp;Mohsen Imani ,&nbsp;Farhad Imani","doi":"10.1016/j.compeleceng.2025.110261","DOIUrl":"10.1016/j.compeleceng.2025.110261","url":null,"abstract":"<div><div>Federated Learning (FL) has become a key method for preserving data privacy in Internet of Things (IoT) environments, as it trains Machine Learning (ML) models locally while transmitting only model updates. Despite this design, FL remains susceptible to threats such as model inversion and membership inference attacks, which can reveal private training data. Differential Privacy (DP) techniques are often introduced to mitigate these risks, but simply injecting DP noise into black-box ML models can compromise accuracy, particularly in dynamic IoT contexts, where continuous, lifelong learning leads to excessive noise accumulation. To address this challenge, we propose Federated HyperDimensional computing with Privacy-preserving (FedHDPrivacy), an eXplainable Artificial Intelligence (XAI) framework that integrates neuro-symbolic computing and DP. Unlike conventional approaches, FedHDPrivacy actively monitors the cumulative noise across learning rounds and adds only the additional noise required to satisfy privacy constraints. In a real-world application for monitoring manufacturing machining processes, FedHDPrivacy maintains high performance while surpassing standard FL frameworks — Federated Averaging (FedAvg), Federated Proximal (FedProx), Federated Normalized Averaging (FedNova), and Federated Optimization (FedOpt) — by up to 37%. Looking ahead, FedHDPrivacy offers a promising avenue for further enhancements, such as incorporating multimodal data fusion.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110261"},"PeriodicalIF":4.0,"publicationDate":"2025-03-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143644963","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":"TFMSNet: A time series forecasting framework with time–frequency analysis and multi-scale processing","authors":"Xin Song ,&nbsp;Xianglong Zhang ,&nbsp;Wang Tian ,&nbsp;Qiqi Zhu","doi":"10.1016/j.compeleceng.2025.110260","DOIUrl":"10.1016/j.compeleceng.2025.110260","url":null,"abstract":"<div><div>Time series forecasting is crucial in various fields. When dealing with complex time series data, existing methods often focus on a single scale or overlook frequency domain information, leading to the loss of critical information. To address this, this paper proposes TFMSNet, a novel time series forecasting framework combining time–frequency analysis with multi-scale processing. The framework decomposes the data into seasonal and trend components. For the seasonal component, TFMSNet utilizes Discrete Wavelet Transform (DWT) to decompose the data into subsequences of different frequencies, combining this with patch-based encoding layers and Inverse DWT to finely capture and reconstruct time–frequency features. It then performs multi-scale analysis and forecasting. For the trend component, the framework achieves multi-resolution representations through downsampling and uses Multilayer Perceptrons (MLPs) for prediction. By integrating both frequency and time domain information and leveraging the multi-scale characteristics of the data, TFMSNet significantly enhances prediction accuracy and robustness. Across 70 results from seven datasets, TFMSNet achieves 48 best and 20 second-best results, demonstrating the best overall performance.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110260"},"PeriodicalIF":4.0,"publicationDate":"2025-03-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143644961","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":"COLO: Combined osprey and lyrebird optimization for optimal antenna selection for massive MIMO system","authors":"Raghunath Mandipudi,&nbsp;Chandra Shekhar Kotikalapudi","doi":"10.1016/j.compeleceng.2025.110245","DOIUrl":"10.1016/j.compeleceng.2025.110245","url":null,"abstract":"<div><div>Massive MIMO (M-MIMO) design is essential for enhancing spatial multiplexing gains in modern communication systems, but it often compromises energy efficiency (EE). Selecting the optimal antenna subset is crucial for boosting EE without negatively impacting spectrum efficiency (SE). However, due to the exponential increase in processing time as antenna count rises, exhaustive search methods become impractical for large MIMO systems. To address this, a novel optimization approach for optimal antenna selection (OAS) is proposed, combining the Osprey Optimization Algorithm (OOA) and Lyrebird Optimization Algorithm (LOA) into a hybrid COLO algorithm. COLO introduces key innovations, including a Feature Dependency-based (FDB) selection technique, a Global Positioning Strategy (GPS) for better search guidance, and OOA integration for enhanced exploration and exploitation. This approach aims to maximize SE while improving system efficiency. The suggested COLO for the maximal scenario has a lower fitness of 4.55×10–09, whereas the traditional LOA, RPO, OOA, EHO, COA, CB-PSO, and GA+CSO+PSO models achieve a higher fitness than COLO. The performance of COLO-based OAS is evaluated against existing methods in terms of efficiency, antenna count, statistical analysis, and convergence, demonstrating its superiority in maximizing SE.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110245"},"PeriodicalIF":4.0,"publicationDate":"2025-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143636272","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 multi-perturbation consistency framework for semi-supervised person re-identification","authors":"Xinyuan Chen ,&nbsp;Yi Niu ,&nbsp;Mingwen Shao ,&nbsp;Weikuan Jia","doi":"10.1016/j.compeleceng.2025.110246","DOIUrl":"10.1016/j.compeleceng.2025.110246","url":null,"abstract":"<div><div>The semi-supervised person re-identification(Re-ID) task only manually annotates a small portion of person identities to reduce costs, but existing methods suffer from insufficient and incomplete utilization of hard unlabeled data, which leads to performance bottleneck. In this paper, we propose a new semi-supervised Re-ID framework to address this issue. In this framework, hard unlabeled samples participate in dual feature consistency learning by generating Multi-perturbation views. The proposed multi-perturbations include three different image-level perturbations and one feature-level perturbation, and the combination of these perturbations can fully simulate the complex changes of persons. To further improve the disturbance quality, a semi-supervised image generation network Semi-DGNet and a Perturbation Scheme Generator (PSG) are proposed to enhance the disturbance effect and control the disturbance intensity. Furthermore, a new Quintuplet loss is proposed to further reduce intra-class distance and increase inter-class distance through a metric learning strategy that involves the joint participation of labeled and unlabeled samples. The above work effectively explores the guiding role of labeled samples in training hard unlabeled data, which has inspiring value for future weakly supervised learning research. Extensive experiments on two datasets and sufficient comparisons with other existing state-of-art methods validate the effectiveness of the proposed framework, and verify its successful integration of multiple training strategies and process, modules, and optimization techniques.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110246"},"PeriodicalIF":4.0,"publicationDate":"2025-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143636273","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 subset selection for big data via parallel chaotic binary differential evolution and feature-level elitism","authors":"Yelleti Vivek ,&nbsp;Vadlamani Ravi ,&nbsp;P. Radha Krishna","doi":"10.1016/j.compeleceng.2025.110232","DOIUrl":"10.1016/j.compeleceng.2025.110232","url":null,"abstract":"<div><div>Feature subset selection (FSS) employing a wrapper approach is fundamentally a combinatorial optimization problem maximizing the area under the receiver operating characteristic curve (AUC) of a classifier built on this subset under single objective environment. To balance both the AUC and the cardinality of the selected feature subset, we propose a novel multiplicative fitness function that combines AUC and a decreasing function of cardinality. Although the differential evolution algorithm is robust, it is prone to premature convergence, which can result in entrapment in local optima. To address this challenge, we propose chaotic binary differential evolution coupled with feature-level elitism (CE-BDE), where the chaotic maps are introduced at the <em>initialization</em> and the <em>crossover</em> operator. We also introduce feature-level elitism to improve the exploitation capability. Feature-level elitism involves preserving those features, which are chosen based on their frequency of occurrence in the population in the evolution process. Dealing with big data entails computational complexity, which motivates us to propose an effective parallel/ distributed strategy island model. The results demonstrate that the parallel CE-BDE outperformed the rest of the algorithms in terms of <em>mean AUC</em> and <em>cardinality</em>. The speedup and computational gain yielded by the proposed parallel approach further accentuate its superiority. Overall, the top-performing algorithm with the multiplicative fitness function turned out to be statistically significant compared to that with the additive fitness function across 5 out of 6 datasets.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110232"},"PeriodicalIF":4.0,"publicationDate":"2025-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143628075","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":"AI-DeepFrothNet: Continuous monitoring and tracking of froth flotation working condition by root cause analysis and optimized predictive control","authors":"Khalid A. Abouda ,&nbsp;Degang Xu ,&nbsp;Wail M. Idress ,&nbsp;Hager M. Elmaki ,&nbsp;Tehseen Mazhar ,&nbsp;Muhammad Aoun ,&nbsp;Yazeed Yasin Ghadi ,&nbsp;Tariq Shahzad ,&nbsp;Habib Hamam","doi":"10.1016/j.compeleceng.2025.110251","DOIUrl":"10.1016/j.compeleceng.2025.110251","url":null,"abstract":"<div><div>Achieving optimal working conditions in froth flotation is critical for maximizing mineral recovery. Traditional manual observation methods are limited by subjectivity and the inability to adapt to changing production environments. Many existing approaches do not provide a clear picture of the flotation behavior's root cause, which directly impacts the grade recovery rate. In this study, we proposed an AI-DeepFrothNet solution to address the prevailing challenges. The proposed work utilized a Putrefaction Enrichment and Tuning Network (PETNET) to eliminate and adjust the noise in the Red, Green, and Blue (RGB) images. Using a Skipped Attention Gated Recurrent Unit (SkA-GRU) for RGB to Hyper Spectral Image (HSI) conversion ensured the preservation of the local and global features. The pre-processed frames were subjected to frame-by-frame analysis using the You Look Only Once-V7 (YOLO-V7). To identify a root cause, the proposed research utilized a Multi-Agent Deep Q Learning (MA-DQL) solution, in which three agents were involved in analyzing the different conditions and properties of the froth layer. To ensure the quality and stability of the mineral outcome, the optimized controller comprehended the root cause control variables and optimized their values using the Gazelle Optimization Algorithm (GOA) logic. The proposed work demonstrated superior performance compared to existing methods and achieved 93 % accuracy, 96 % precision, 95 % recall, and 87 % F1 score, outperforming other methods.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110251"},"PeriodicalIF":4.0,"publicationDate":"2025-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143628074","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":"Impact of phase-locked loop on grid-connected inverter stability under weak grid conditions and suppression measures","authors":"Yueyang Zheng ,&nbsp;Yang Han ,&nbsp;Congling Wang ,&nbsp;Quan Ren ,&nbsp;Ping Yang ,&nbsp;Amr S. Zalhaf","doi":"10.1016/j.compeleceng.2025.110249","DOIUrl":"10.1016/j.compeleceng.2025.110249","url":null,"abstract":"<div><div>The growing portion of renewable energy in the energy mix has led to the gradual emergence of weak or very weak grid characteristics with high impedance. In this context, the phase-locked loop (PLL) and its interaction with other key control links present a significant challenge to the stable operation of grid-connected inverters. Recent studies have focused on PLL induced frequency coupling and negative impedance characteristics and their impact on system stability. However, there is a lack of comprehensive compilation and systematic summarization of these results, making subsequent research direction unclear. This paper comprehensively summarizes the existing literature and concludes that the structure of the Phase-Locked Loop (PLL) leads to frequency coupling within the system, potentially inducing harmonic oscillations. Specifically, when the PLL bandwidth is excessively wide, it enhances the dynamic response of the system, simultaneously broadening the range of influence where negative damping phenomena occur. Conversely, when the PLL bandwidth is overly narrow, the stability of the inverter is improved, albeit at the cost of compromised dynamic performance. Additionally, the paper examines the frequency coupling phenomenon generated by PLL and its negative impedance characteristics. Based on the analysis, the paper systematically summarizes and discusses methods to enhance system robustness through PLL parameter adjustment, filter design, and voltage feedforward control. Furthermore, it considers the PLL as an active disturbance channel in the grid system and explores how other potential disturbances affect overall system stability through the PLL. Lastly, the article highlights the shortcomings of current research, identifies key points and challenges, and provides valuable references for future research directions.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110249"},"PeriodicalIF":4.0,"publicationDate":"2025-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143628076","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":"Integrating frequency limitation and feature refinement for robust 3D Gaussian segmentation","authors":"Xinyu Hu ,&nbsp;Haijian Shao ,&nbsp;Xing Deng ,&nbsp;Yingtao Jiang ,&nbsp;Fei Wang","doi":"10.1016/j.compeleceng.2025.110239","DOIUrl":"10.1016/j.compeleceng.2025.110239","url":null,"abstract":"<div><div>3D scene segmentation is a core challenge in computer vision, aiming to precisely and efficiently extract targets from complex 3D environments. Despite some research progress, existing methods struggle with high-frequency information in complex, multi-scale and multi-view scenarios, facing issues like hard-to-track high-frequency data, high resource consumption, low segmentation accuracy, and complex user interactions, which limit the technology’s practical use and development. To address the above problems, this paper proposes a robust 3D Gaussian segmentation method (IFFSG) that integrates frequency restriction and feature refinement. By innovating a frequency-adaptive sampling constraint strategy, a 3D frequency modulation filter is constructed, dynamically adjusting the frequency of the 3D Gaussian elements according to multi-view input constraints. The highest frequency of the reconstructed 3D Gaussian scene is strictly confined within the sampling frequency range of the input views. Accurately matching the actual frequency and sampling frequency of the scene, effectively avoiding artifacts caused by scale changes, and significantly improving the accuracy and quality of scene reconstruction in terms of geometric structure and texture details. This provides accurate and stable basic data for aligning 2D and 3D features in subsequent segmentation. Additionally, this paper introduces a robust multi-view feature alignment strategy, using the advanced segmentation capability of SAM to guide the training of 3D features and promote the alignment of similar features in 3D space. This strategy promotes close alignment of similar features in 3D space, greatly enhancing the compactness and consistency of the features. During segmentation, the model relies on this optimized feature to more sensitively and accurately identify object boundaries and structures, providing fine and reliable data support for subsequent scene understanding and analysis. To validate the effectiveness of the proposed method, experiments are conducted on several challenging datasets. The results show that this method achieves near-real-time segmentation speeds with minimal user input, significantly outperforming existing techniques. On the NVOS dataset, its accuracy reaches SOTA level while maintaining a near-real-time inference time of 0.2 s. In terms of image quality assessment, we obtained SSIM of 0.801, PSNR of 20.04, and LPIPS of 0.173.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110239"},"PeriodicalIF":4.0,"publicationDate":"2025-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143627974","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":"RGB-Thermal cameras calibration based on Maximum Index Map","authors":"Jiahui Wei ,&nbsp;Zhongwen Zou ,&nbsp;Wenjie Lai ,&nbsp;Junliang Du ,&nbsp;Jiaxue Zhao ,&nbsp;Ziji Liu ,&nbsp;Shengzhe Wang","doi":"10.1016/j.compeleceng.2025.110234","DOIUrl":"10.1016/j.compeleceng.2025.110234","url":null,"abstract":"<div><div>Calibration across multiple sensors is crucial in computer vision tasks, such as image registration, 3D reconstruction, and target tracking. RGB-Thermal (RGB-T) cameras, widely used as a multi-sensor combination, traditionally depend on manual calibration methods to determine extrinsic parameters. However, this manual process is intricate and lacks the capability for online calibration during use. To overcome these challenges, an online calibration algorithm for RGB-T camera extrinsic parameters is proposed. The algorithm first addresses modal differences by matching feature point pairs using the Maximum Index Map (MIM) feature. These matched features are then used to calculate the extrinsic parameters through homography and epipolar constraints between images. Additionally, a convenient intrinsic calibration method is introduced, one that does not rely on extrinsic infrared light sources, thereby overcoming the limitations of standard calibration boards, which are often unsuitable for RGB-T cameras. Experimental results demonstrate that the proposed online extrinsic calibration algorithm significantly simplifies the calibration process compared to traditional methods, achieving accurate calibration with only a pair of images. The method achieves an average reprojection error of 0.677 pixels for RGB images and 0.635 pixels for thermal images, highlighting its precision. The method’s effectiveness is further validated by the precision in RGB-T image registration.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110234"},"PeriodicalIF":4.0,"publicationDate":"2025-03-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143621015","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}