Computers & Electrical Engineering最新文献

{"title":"Advancing harmonic prediction for offshore wind farms using synthetic data and machine learning","authors":"Alp Karadeniz","doi":"10.1016/j.compeleceng.2025.110613","DOIUrl":"10.1016/j.compeleceng.2025.110613","url":null,"abstract":"<div><div>This study presents a novel forecasting model for accurate harmonic prediction in offshore wind farms (OWFs) using data augmentation and machine learning techniques. A Generative Adversarial Network (GAN) is employed to generate synthetic meteorological data, enhancing the training set for improved accuracy. The model utilizes wind speed data from Bozcaada, Turkey, and simulates voltage and current waveforms to predict Total Harmonic Distortion Voltage (THDV). Machine learning (Random Forest) and deep learning (LSTM, GRU) models are compared to assess prediction performance. Results show that the GAN-based data augmentation significantly enhances prediction accuracy. This study provides a valuable methodology for harmonic forecasting in OWFs, offering insights for future renewable energy system planning and grid stability.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110613"},"PeriodicalIF":4.9,"publicationDate":"2025-08-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144771698","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 three-stage framework for multi-type pantograph anomaly detection under complex environments","authors":"Hao Yan ,&nbsp;Chuan Lin ,&nbsp;Ningning Guo ,&nbsp;Zhiyuan Xu ,&nbsp;Jiefeng Zang ,&nbsp;Anyong Qing","doi":"10.1016/j.compeleceng.2025.110612","DOIUrl":"10.1016/j.compeleceng.2025.110612","url":null,"abstract":"<div><div>A novel three-stage framework is proposed in this paper for detecting pantograph anomalies. This framework is capable of detecting anomalies in multiple types of pantographs and is resilient to complex backgrounds and illumination variations, exhibiting strong robustness. In the first stage, the improved Yolov8 network is utilized to localize the pantograph region, addressing the issue of complex backgrounds during pantograph detection. In the second stage, the Short-Term Dense Concatenate (STDC) network is employed for precise segmentation of the pantograph region. Furthermore, corresponding improvements are made to the network to handle edge blurring caused by illumination variations. In the third stage, binary images of different types of pantographs are transformed into vectors that contain pantograph features. Additionally, Relief-F and random forest algorithms are employed for feature selection and anomaly classification. Ultimately, the proposed framework achieves an average accuracy of 97.04% for various anomaly types in a testing set consisting of images of multiple types of pantographs.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110612"},"PeriodicalIF":4.9,"publicationDate":"2025-08-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144771334","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":"Evolutionary computation based wind energy integrated multi-objective optimal reactive power dispatch and economic load dispatch problem","authors":"Tanmay Das ,&nbsp;Ranjit Roy ,&nbsp;Kamal Krishna Mandal","doi":"10.1016/j.compeleceng.2025.110587","DOIUrl":"10.1016/j.compeleceng.2025.110587","url":null,"abstract":"<div><div>The study explores power loss and voltage fluctuations in transmission networks, integrating them with the economic load dispatch problem (ELD) to create a new multi-objective optimization framework, aiming to minimize these issues while reducing generation costs. The multi-objective function combines three goals: minimizing ELD-related generation costs, power loss, and total voltage deviation (TVD) associated with optimal reactive power dispatch (ORPD). The framework also includes a renewable energy source (RES) optimally positioned to assess its impact on the system and the specific objectives of the multi-objective (MO)-ELD-ORPD problem. The variability of wind energy is considered, with optimal wind power output determined using the Weibull distribution function (WPDF). The analysis is conducted for the IEEE 30 bus system using the JAYA algorithm, while a fuzzy-based Pareto optimality method identifies the best solutions. Additionally, a special case for the larger IEEE 118 bus system is examined, accounting for uncertain power contributions from photovoltaic (PV) and wind sources, using a realistic RES model based on Srinagar, India. The JAYA algorithm has successfully provided optimal solutions for the MO-ELD-ORPD, resulting in significant reductions in generation costs, power losses, and TVD compared to other evaluated algorithms. Utilizing overestimated wind power led to decreases of approximately 4.97 %, 23.7 %, and 31.38 % in generation costs, power losses, and TVD, respectively, in the IEEE 30 bus system. Additionally, incorporating PV and wind energy in the IEEE 118 bus system significantly improved the proposed case study results compared to scenarios without RES, thus validating the effectiveness of this method.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110587"},"PeriodicalIF":4.9,"publicationDate":"2025-08-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144766418","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 skill one-to-one-based optimization enabled trajectory planning in Internet of Things","authors":"Anand R. Umarji ,&nbsp;Dharamendra Chouhan","doi":"10.1016/j.compeleceng.2025.110605","DOIUrl":"10.1016/j.compeleceng.2025.110605","url":null,"abstract":"<div><div>With the advancement of communication technology, unmanned aerial vehicles (UAVs) have been extensively utilized for attaining prolonged exposure to the Internet of Things (IoT). Due to the high sensitivity nature of sudden path changes, obstacle interference, and limited adaptability to dynamic environments, the performance of the UAV trajectory remains poor. To solve the above issues, this research proposed a hybrid optimization model called Skill One-to-One-Based Optimization (SOOBO) to initially generate an optimal, constraint-aware trajectory. This model employed a trajectory correction mechanism to adjust the path to avoid potential collision. Initially, the UAV-IoT model is taken into account, with trajectory generation incorporating both range constraints and collision avoidance among UAVs. The proposed SOOBO is employed for generating the feasible trajectory. Here, the SOOBO is obtained by the integration of a Skill Optimization Algorithm (SOA) and a One-to-One-Based Optimizer (OOBO). OOBO is a metaheuristic approach that effectively resolves optimization issues through iterative processes. The OOBO effectively solves the optimization issues and provides effectual quasi-optimal solutions. To attain a more effectual solution with faster convergence speed, the SOA is added to OOBO. SOA is based on the inspiration from an individual’s desire for learning and improving their knowledge. The SOA covers two stages termed as exploration and exploitation. Moreover, trajectory correction is performed to avoid collision between UAVs and obstacles. For attaining a better trajectory, the inscribed circle (IC) smooth method is utilized. Moreover, the performance measuring parameters like path length, speed, residual energy, and fitness are considered to estimate the performance of SOOBO-based trajectory planning in IoT, in which the finest outcomes of 12.54, 20.97m/s, 0.412 J, and 0.792 are attained.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110605"},"PeriodicalIF":4.9,"publicationDate":"2025-08-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144766419","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":"Techno-economic performance assessment of large-scale distribution network supported by renewable sources and compressed air energy storages considering uncertainties and demand response","authors":"Ahmed T. Hachemi ,&nbsp;Mohamed Hashem ,&nbsp;Abdelhakim Saim ,&nbsp;Imen Ben Hamida ,&nbsp;Medhat E.M. Ali ,&nbsp;Francisco Jurado ,&nbsp;Mohamed Ebeed","doi":"10.1016/j.compeleceng.2025.110600","DOIUrl":"10.1016/j.compeleceng.2025.110600","url":null,"abstract":"<div><div>The increasing integration of renewable distributed generations (RDGs) has made energy management (EM) in distribution networks (DNs), along with energy storage systems (ESSs), critically important. This study proposes a probabilistic EM framework for the techno-economic assessment of large-scale DNs, focusing on voltage stability, voltage profile, and total annual cost. The framework optimally allocates and operates wind turbines (WTs), photovoltaic units (PVs), and compressed air energy storage systems (CAESs), while incorporating real-time pricing (RTP)-based demand side response (DSR) and uncertainties. These uncertainties related to solar irradiance, temperature, wind speed, load demand, and energy price are modeled using Monte Carlo Simulation. A novel Adaptive Chernobyl Disaster Optimizer (ACDO) is developed and employed to solve the multi-objective optimization problem. The robustness of ACDO is evaluated against other metaheuristic algorithms using ten benchmark functions and the Friedman test. The proposed EM strategy is validated on the IEEE 118-bus DN through three case studies: (i) RDGs only, (ii) RDGs with CAESs, and (iii) RDGs with CAESs and DSR. Results from case (iii) show an 87.1 % cost reduction, a 6.95 % improvement in voltage stability, and a 54.9 % reduction in voltage deviations. These outcomes confirm the effectiveness of the proposed ACDO-based EM in enhancing DN performance.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110600"},"PeriodicalIF":4.9,"publicationDate":"2025-08-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144766917","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":"ChAT-BiGRU-NBEATS: An efficient and robust deep learning model for time series weather data prediction","authors":"Naba Krushna Sabat ,&nbsp;Rashmiranjan Nayak ,&nbsp;Umesh Chandra Pati ,&nbsp;Santos Kumar Das","doi":"10.1016/j.compeleceng.2025.110603","DOIUrl":"10.1016/j.compeleceng.2025.110603","url":null,"abstract":"<div><div>Data-driven forecasting models are used to understand environmental climatology data better but often overlook missing values and noise, leading to ineffective temporal modeling and inadequate correlations between weather parameters. Consequently, these limitations adversely impact the accuracy of predictions. In response to these issues, a novel deep hybrid model, i.e., a channel attention-enabled bidirectional gated recurrent unit with neural basis expansion analysis for time series (ChAT-BiGRU-NBEATS), is proposed. In this case, a bidirectional gated recurrent unit (Bi-GRU) network is augmented with a channel-attention mechanism and an NBEATS model that facilitates the extraction of complex data features and the prediction of long data sequences. The efficacy of the proposed model is assessed using a comparative analysis against several state-of-the-art deep learning models, utilizing error metrics such as mean absolute error (MAE), mean squared error (MSE), root mean squared error (RMSE), and the <span><math><msup><mrow><mi>R</mi></mrow><mrow><mn>2</mn></mrow></msup></math></span> score. It is evident from the results that the proposed hybrid model surpasses other models in terms of its heightened accuracy.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110603"},"PeriodicalIF":4.9,"publicationDate":"2025-08-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144766417","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":"Localization based on received signal strength measurement in smart cities using constrained least squares","authors":"Bamrung Tausiesakul ,&nbsp;Emanuele Goldoni ,&nbsp;Pietro Savazzi","doi":"10.1016/j.compeleceng.2025.110564","DOIUrl":"10.1016/j.compeleceng.2025.110564","url":null,"abstract":"<div><div>The global positioning system (GPS) is essential for many internet-of-things applications but is vulnerable to spoofing and jamming attacks that can lead to incorrect location and timing information. This paper proposes a GPS-compromise detection and localization method using received signal strength (RSS) from wireless networks as a low-cost alternative. Although RSS measurements are inherently noisy, they can provide useful location estimates when processed effectively. We formulate a localization problem using noisy RSS data and propose three estimation methods based on a constrained least squares (CLS) criterion. The Cramér–Rao lower bound for mean squared error is also derived to evaluate the performance limits. Simulations based on real-world LoRaWAN data show that the proposed CLS methods achieve lower estimation error, measured by root mean squared error, than the conventional least squares method, albeit at a higher computational cost.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110564"},"PeriodicalIF":4.9,"publicationDate":"2025-08-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144757835","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":"Multidimensional CapsNets attention-gated approach for skin cancer detection and classification","authors":"Sonali R Nalamwar ,&nbsp;Sandeep U. Belgamwar","doi":"10.1016/j.compeleceng.2025.110573","DOIUrl":"10.1016/j.compeleceng.2025.110573","url":null,"abstract":"<div><div>Skin cancer remains a major global cause of mortality, and early detection in its premalignant stages is crucial for improving patient outcomes. Traditional diagnostic methods face challenges such as time-consuming analysis, and limited accuracy. This study introduces the Multidimensional Capsule Networks Attention-Gated Module (MCAGM), an advanced automated deep learning framework designed to overcome these limitations. The MCAGM model utilizes Capsule Networks (CapsNets) enhanced with an pioneering spatial-channel attention mechanism, specifically designed to highlight clinically significant features in dermoscopic images (HAM10000 dataset) while effectively suppressing noise. The dual-domain attention mechanism (spatial and channel) dynamically refines feature importance, eliminating subjective interpretation and ensuring objective prioritization of relevant features. This end-to-end automated system dramatically reduces diagnosis time from hours to seconds, offering a significant improvement in efficiency. Furthermore, the CapsNet-based spatial hierarchies preserve critical lesion patterns that are often missed by conventional Convolutional Neural Networks (CNNs), enhancing the model's ability to detect subtle features and improve diagnostic accuracy. The model achieves exceptional performance with 97.63 % accuracy, 98.11 % precision, and 98.73 % recall, outperforming state-of-the-art methods by 8–19 % in accuracy (e.g., CNN: 88.88 %, CapsNet: 86.84 %), demonstrating its potential as a reliable tool for skin cancer diagnosis.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110573"},"PeriodicalIF":4.9,"publicationDate":"2025-08-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144757836","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":"CWGCN: Cascaded Wavelet Graph Convolution Network for pedestrian trajectory prediction","authors":"Wangxing Chen,&nbsp;Haifeng Sang,&nbsp;Zishan Zhao","doi":"10.1016/j.compeleceng.2025.110609","DOIUrl":"10.1016/j.compeleceng.2025.110609","url":null,"abstract":"<div><div>Accurately predicting the future trajectory of pedestrians is crucial for the practical application of autonomous driving, service robots, and intelligent monitoring systems. Previous methods have ignored capturing interaction features at multiple scales. Additionally, these methods directly add pedestrian social and temporal interaction features, resulting in one type of feature dominating the model prediction results and thus affecting the overall prediction accuracy. To address these problems, we propose a cascaded wavelet graph convolution network. Specifically, the network first constructs spatial and temporal graphs and employs a self-attention mechanism to obtain an attention score matrix that preliminarily describes the social and temporal interaction relationships of pedestrians. Next, we design a cascaded wavelet transform module to process the attention score matrix, capturing multi-scale interaction features through cascaded wavelet transforms and asymmetric convolution. We then create a spatial–temporal guided fusion module to achieve reasonable weighting of pedestrian social and temporal interaction features through spatial–temporal guided attention. Finally, we utilize temporal convolutional networks to predict multiple future trajectories directly. Experiments on three public datasets demonstrate that our method exhibits excellent prediction performance and is more effective in pedestrian instantaneous and long-term trajectory prediction.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110609"},"PeriodicalIF":4.9,"publicationDate":"2025-08-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144757837","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":"Deep Guard: An enhanced hybrid ensemble classifier for face presentation attack detection integrating Gabor and binarized statistical image features descriptors with deep learning","authors":"Aparna Santra Biswas ,&nbsp;Somnath Dey ,&nbsp;Sanskar Verma ,&nbsp;Khushi Verma","doi":"10.1016/j.compeleceng.2025.110566","DOIUrl":"10.1016/j.compeleceng.2025.110566","url":null,"abstract":"<div><div>Facial recognition systems are widely used in various real-world applications due to their reliability and convenience. However, attackers exploit these systems by mimicking bona fide user traits to gain unauthorized access. This emphasizes the need for effective countermeasures to be integrated into face-based authentication systems. Face presentation attack detection methods encounter several challenges such as illumination variations and noisy input images which limit the performance of the attack detection methods, particularly on unseen data. In this paper, we introduce Deep Guard, a hybrid framework that combines handcrafted texture descriptors with advanced deep learning techniques. The framework utilizes an ensemble of different classifiers to leverage their complementary strengths. The first classifier applies Binarized Statistical Image Features (BSIF) and a Multilayer Perceptron (MLP) to capture fine-grained texture details. The second classifier combines EfficientNet-B0 with ConvMixer layers and a CBAM attention mechanism to enhance feature representation and improve perceptual capabilities. The third classifier uses Gabor filters as convolutional layers with a deep network which is used in second classifier to refine edges and increase robustness to illumination and noise. The outputs from these classifiers are fused using a soft voting mechanism to classify facial images as real or fake. We evaluate the proposed framework on six publicly available datasets CASIA-FASD, Replay-Attack, 3DMAD, ROSE-Youtu, OULU-NPU, and MSU-MFSD. The results demonstrate that Deep Guard outperforms most state-of-the-art methods in intra-dataset testing and achieves strong generalization performance in cross-dataset single source training and testing scenarios, with an average HTER of 25.78% for HybridNet I, which combines all three classifiers and 27.96% for HybridNet II, combining classifiers two and three. It also achieves an AUC of 98.65% for cross-dataset evaluation with multiple-source training and single-source testing (O&amp;C&amp;I <span><math><mo>→</mo></math></span> M).</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110566"},"PeriodicalIF":4.9,"publicationDate":"2025-07-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144738098","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}