Computers & Electrical Engineering最新文献

{"title":"Pareto front-based multi-objective optimization of renewable DGs in distribution networks using an enhanced Newton-Raphson algorithm with techno-economic and environmental assessment","authors":"Mohamed A. Elseify ,&nbsp;Salah Kamel","doi":"10.1016/j.compeleceng.2025.110637","DOIUrl":"10.1016/j.compeleceng.2025.110637","url":null,"abstract":"<div><div>Distributed generation (DG) is mostly installed at the distribution level of the power grid. The advantages of DG systems are crucial in the enlargement and design of the overall power system. The analysis of DG allocation in a distribution network (DN) necessitates the achievement of three primary objectives: minimizing the investment cost of DG, decreasing greenhouse gas (GHG) emissions, and improving the technical performance of the DNs. In this paper, these multiple goals are formulated as single/multi-objective functions and addressed using a modified Newton-Raphson-based optimization (mNRBO). The Weibull probability operator and leader-based mutation techniques are initially added to the basic NRBO to improve its performance. The single objective seeks to maximize the net profit due to the reduced purchasing energy from the sub-stations and the reduced penalty cost of GHG emissions. The multiple goals aim to optimize the most conflicting objectives: total active power loss and total DG cost. This multi-objective is based on the Pareto-based archiving technique and the best compromise solution (BCS) is chosen using the fuzzy decision-based-making algorithm. Initially, the robustness of the multi-objective mNRBO (MOmNRBO) is evaluated using diverse test benchmark functions with various forms of Pareto optimal front (POF), such as ZDT, MMF, and UF, using multiple performance metrics, including spacing, inverse generation distance (IGD), and generation distance (GD). Then, the proposed technique is applied to properly integrate different types of DG in the real Egyptian 37-bus, the 69-bus, and the 118-bus DNs. The superiority and efficiency of the mNRBO are also demonstrated over the original NRBO and other boosted algorithms reported in the literature, considering different statistical analyses. The results revealed that the proposed algorithm is better in achieving global solutions in different situations, maximizing the techno-economic, and environmental advantages of practical and standard DNs.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"128 ","pages":"Article 110637"},"PeriodicalIF":4.9,"publicationDate":"2025-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144907486","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":"Leveraging large language models for sentiment analysis: A hybrid contextual and linguistic feature fusion approach","authors":"Le An Pha ,&nbsp;Thien Khai Tran","doi":"10.1016/j.compeleceng.2025.110643","DOIUrl":"10.1016/j.compeleceng.2025.110643","url":null,"abstract":"<div><div>Sentiment classification remains a complex natural language processing task due to implicit cues and contextual variability, especially in multilingual or low-resource scenarios. This paper proposes CL-BERT, a hybrid framework that integrates large language models with BERT through chain-of-thought prompting. By generating linguistic and contextual explanations, CL-BERT enhances semantic understanding and improves classification performance. These explanations are fused into BERT representations using a learnable attention mechanism. Experiments on Vietnamese, English, and Bambara datasets show that CL-BERT outperforms strong baselines, demonstrating its robustness and adaptability. The model offers a practical and interpretable solution for sentiment analysis across diverse languages and domains.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110643"},"PeriodicalIF":4.9,"publicationDate":"2025-08-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144892022","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 novel intra-layer mixed bit-width quantization method for the classification of 1D periodic time-series signals","authors":"Zijing Xia ,&nbsp;Huaiyu Zhu ,&nbsp;Haitao Ying ,&nbsp;Junjie Li ,&nbsp;Ruohong Huan ,&nbsp;Yun Pan","doi":"10.1016/j.compeleceng.2025.110633","DOIUrl":"10.1016/j.compeleceng.2025.110633","url":null,"abstract":"<div><div>Deep learning (DL) methods have been successfully applied to classification tasks involving one-dimensional periodic time-series signals, such as mechanical vibration, electrocardiograms (ECGs) and photoplethysmograms (PPG). Despite notable success, mainstream DL models face challenges in adapting to diverse mobile scenarios. To tackle these issues, quantization techniques provide an effective approach to reduce model size and improve inference speed, facilitating easier deployment on mobile devices. In this study, we introduced an Intra-layer Mixed Bit-width Quantization (IMBQ) method that leverages signals’ periodicity to address existing limitations. We developed optimal feature fragment splitting strategies for each layer and validated the optimal bit-width allocation through both manual traversal and regularization methods, thereby highlighting critical information within feature layers. Results demonstrated that our proposed method achieved superior model compression ratios and inference speeds while maintaining top-tier performance, surpassing current state-of-the-art solutions. This research offered new insights and approaches for analyzing such signals in mobile environments.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110633"},"PeriodicalIF":4.9,"publicationDate":"2025-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144892021","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":"GymProxy: A lightweight Python wrapper for seamless integration of standalone simulators into Gymnasium environments","authors":"Sae Hyong Park ,&nbsp;Seungjae Shin ,&nbsp;Namseok Ko ,&nbsp;Taeyeon Kim ,&nbsp;Dongman Lee","doi":"10.1016/j.compeleceng.2025.110634","DOIUrl":"10.1016/j.compeleceng.2025.110634","url":null,"abstract":"<div><div>Reinforcement learning (RL) often relies on extensive simulation-based training to refine agent behavior, as real-world errors can have significant repercussions. Integrating agents into existing standalone simulators typically requires code modifications and careful implementation of the synchronization mechanisms that ensure the agent and simulator execute in a turn-taking manner with proper data exchanges, which is often delicate and error-prone. In this paper, we introduce GymProxy, a lightweight and user-friendly open-source wrapper library designed to simplify the integration of standalone simulators into Gymnasium, the leading framework for developing RL environments. GymProxy provides simple and abstracted application programming interfaces (APIs) for seamlessly integrating an external simulator with the Gymnasium environment while handling synchronization internally. With GymProxy, developers can quickly build a custom Gymnasium environment that wraps existing standalone simulators with minimal implementation effort, eliminating the need to manually program complex synchronization logic.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110634"},"PeriodicalIF":4.9,"publicationDate":"2025-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144889512","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":"Distilling Large Language Models for Structured Query Language generation with reverse Kullback–Leibler and dynamic α scheduler","authors":"Nhat Le ,&nbsp;Quan Ninh ,&nbsp;Tung Le ,&nbsp;Huy Tien Nguyen","doi":"10.1016/j.compeleceng.2025.110607","DOIUrl":"10.1016/j.compeleceng.2025.110607","url":null,"abstract":"<div><div>With the increasing reliance on data-driven decision-making, the demand for efficient Structured Query Language (SQL) query generation has grown significantly, as it serves as a crucial bridge between natural language and databases. While Large Language Models (LLMs) excel in this task, their high computational costs and inconsistent effectiveness pose significant limitations. This study introduces a knowledge distillation approach to create efficient, high-performing models for SQL generation. By integrating teacher and student model distributions with a dynamic <span><math><mi>α</mi></math></span> scheduler inspired by learning rate schedulers, the method adjusts the teacher’s influence during training, enhancing stability and narrowing performance gaps. Additionally, reverse Kullback–Leibler Divergence (KLD) loss balances contributions, allowing the student model to refine itself while leveraging teacher guidance. The resulting distilled student model, which is 100 times smaller than GPT-4, achieves 80.5% accuracy on benchmark datasets and outperforms GPT-4 in this domain. Furthermore, it demonstrates a 10.2% improvement on extra-hard questions compared to its undistilled counterpart. This work highlights the potential of optimizing LLMs for reduced computational costs and superior performance in SQL query generation and beyond.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110607"},"PeriodicalIF":4.9,"publicationDate":"2025-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144885890","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":"Unveiling Android security testing: A Comprehensive overview of techniques, challenges, and mitigation strategies","authors":"Durga Viswanath Palutla ,&nbsp;Sriramulu Bojjagani ,&nbsp;Sai Charan Reddy Mula ,&nbsp;Ravi Uyyala ,&nbsp;Neeraj Kumar Sharma ,&nbsp;Mahesh Kumar Morampudi ,&nbsp;Muhammad Khurram Khan","doi":"10.1016/j.compeleceng.2025.110620","DOIUrl":"10.1016/j.compeleceng.2025.110620","url":null,"abstract":"<div><div>With the rapid growth of Android applications, ensuring robust security has become a critical concern. Traditional Vulnerability Assessment and Penetration Testing (VAPT) approaches, though effective across platforms, often fall short in addressing Android-specific security challenges. This paper presents a comprehensive review of security testing methods tailored to the Android ecosystem, including static and dynamic analysis, hybrid approaches, network communication testing, reverse engineering, malware detection, and permission-based assessments. Android’s open-source nature, device fragmentation, and inconsistent security policies introduce unique vulnerabilities that require specialized testing strategies. By examining current tools, methodologies, and best practices, this review identifies recurring gaps in the Android application security testing process. It highlights the need for more adaptable and thorough testing frameworks. The insights provided are valuable to developers, researchers, and security professionals aiming to strengthen Android app security. Ultimately, this work underscores the importance of tailoring security assessment practices to the evolving threat landscape of the Android platform, thereby contributing to the development of safer and more resilient applications.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110620"},"PeriodicalIF":4.9,"publicationDate":"2025-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144889513","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":"Pool Free Rapid Segmentation Network (PFRS-Net) to detect human blastocyst compartments for embryonic assessment","authors":"Abida Hussain ,&nbsp;Adnan Haider ,&nbsp;Saima Ashraf ,&nbsp;Syed Muhammad Ali Imran ,&nbsp;Muhammad Arsalan","doi":"10.1016/j.compeleceng.2025.110636","DOIUrl":"10.1016/j.compeleceng.2025.110636","url":null,"abstract":"<div><div>Assisted reproductive technology has become an increasingly popular solution to address infertility in humans, primarily by in vitro fertilization (IVF). IVF is a complex process where eggs and sperm are combined outside the human body. This occurs in a controlled, specialized laboratory setting that supports and encourages the growth of embryos before they are transferred to the uterus. The IVF process is carefully monitored and regulated within a laboratory environment until the embryos develop and progress to the blastocyst stage. The standard procedure for in vitro fertilization (IVF) involves transferring one or two blastocysts from a batch that has been developed under controlled conditions. A detailed morphological analysis of these blastocysts is performed, assessing their distinct components, including the trophectoderm (TE), zona pellucida (ZP), inner cell mass (ICM), and blastocoel (BL), using manual microscopic techniques. Although deep learning has been successfully utilized in various medical diagnostic and analytical applications, its integration for automating the morphological analysis of human blastocysts continues to face several obstacles. Current methodologies often exhibit inaccuracies and necessitate considerable preprocessing along with expensive computational architectures. As a result, further research is needed to improve the accuracy and efficiency of deep learning techniques in this field to enable their full potential in assisted reproductive technology. To address this challenge, we introduce the Pool Free Rapid Segmentation Network (PFRS-Net), which is specifically developed to effectively identify the compartments of human blastocysts without relying on pooling operations. The network utilizes rapid convolutional block (RCB) modules to achieve accurate detection. The RCB module is specifically designed to capture valuable deep features with computational efficiency. The Swift Decoder block is then used to up-sample the feature maps to their original size using a few layers. This specialized design helps to reduce the number of trainable parameters while maintaining high segmentation accuracy and recovering the lost spatial information using a feature enhancement block (FEB). Our proposed PFRS-Net accurately detects the blastocyst compartments without preprocessing the image and consuming 1.1 million trainable parameters only. This method is trained and tested using a publicly accessible dataset of human blastocyst images. The experimental outcomes demonstrate superior segmentation performance in detecting blastocyst components, which is vital for embryonic research and analysis.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110636"},"PeriodicalIF":4.9,"publicationDate":"2025-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144889511","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":"Secure and scalable trust management in IoT: A hierarchical blockchain-based approach","authors":"Elham Meybodian ,&nbsp;Seyedakbar Mostafavi ,&nbsp;Tooska Dargahi ,&nbsp;Vahid Ranjbar Bafghi","doi":"10.1016/j.compeleceng.2025.110631","DOIUrl":"10.1016/j.compeleceng.2025.110631","url":null,"abstract":"<div><div>The diverse and resource-constrained nature of Internet of Things (IoT) devices make them vulnerable against various security attacks. Effective trust management within the IoT ecosystem is crucial for reliable data collection and sharing, as well as the detection of malicious nodes. Centralized trust management methods are inefficient due to several challenges, including single point of attack/failure, unauthorized manipulation of trust data, resource limitations of smart devices, and scalability issues. Blockchain technology provides a suitable solution for trust management due to its decentralization, transparency, and immutability features. However, deploying blockchain for IoT devices is not simple due to the low performance and high computational costs of consensus algorithms, limited resources of smart devices, and the large volume of transactions created by nodes. In this paper, a hierarchical trust management approach based on blockchain is proposed. The proposed approach evaluates each node’s reputation and organizational trust at both intra-organizational and inter-organizational levels. At the internal level, a lightweight blockchain is used to evaluate and store the trust score of the nodes. At the inter-organizational level, interactions between organizations and their trust level are recorded in the public blockchain. Two methods are proposed, i.e. probing-based and evidence-based, for evaluating the reputation of each node and the trust level of each organization. The evaluation results show that with a maximum of 35% malicious nodes within an organization, the proposed method can correctly identify the malicious and honest nodes. The recall and specificity measures obtained are both greater than 0.9. Additionally, organizations with more than 35% of malicious nodes are blacklisted and suspended.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110631"},"PeriodicalIF":4.9,"publicationDate":"2025-08-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144879314","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":"An efficient essential secret sharing: Application to gray and color images","authors":"Ramakant Kumar ,&nbsp;Avishek Adhikari ,&nbsp;Sahadeo Padhye ,&nbsp;Mainejar Yadav","doi":"10.1016/j.compeleceng.2025.110630","DOIUrl":"10.1016/j.compeleceng.2025.110630","url":null,"abstract":"<div><div>A Secret Sharing Scheme (SSS) is a cryptographic primitive used to share secrets, such that the secret can be revealed only when a specified number of shares are available. Due to varying levels of importance or hierarchy among shareholders, some shareholders may be deemed essential for secret recovery. This concept is addressed by essential secret sharing. The essential SSSs designed for sharing secret images suffer from various limitations, including different shadow image sizes for essential and non-essential shareholders, large share sizes, preprocessing requirements, the need for concatenating sub-shadows, constraints on image type, and pixel expansion. To address these issues, we propose an essential SSS based on simple linear algebra. This scheme is both perfect and ideal. We then leverage this scheme to develop a perfect and ideal essential secret image sharing (SIS) scheme that accommodates essential participants. We further modify the scheme to reduce the share sizes by a factor of k, which makes it efficient. It is applicable for gray as well as color images. Our approach relies on simple operations like matrix addition, multiplication, and inversion. Our scheme supports color images and offers lossless recovery. It overcomes the limitations present in existing SIS schemes. We have discussed experimental results for the gray and color images. Furthermore, all schemes presented in this article are quantum secure.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110630"},"PeriodicalIF":4.9,"publicationDate":"2025-08-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144879313","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":"Extraction and encryption of facial information for government personnel: STP-CS with Tri-color shuffling scheme","authors":"Linlin Tan ,&nbsp;Suo Gao ,&nbsp;Nanrun Zhou ,&nbsp;Yinghong Cao ,&nbsp;Jun Mou","doi":"10.1016/j.compeleceng.2025.110624","DOIUrl":"10.1016/j.compeleceng.2025.110624","url":null,"abstract":"<div><div>Facial recognition security is a critical concern in the government sector. Protecting individuals’ facial information is essential for ensuring the confidentiality of documents accessed via facial recognition, thereby preventing theft and damage. This paper proposes a facial image protection scheme based on a semi-tensor product compressed sensing (STP-CS) and tri-color shuffling scheme (TSS). First, facial information is extracted and compressed to reduce the data volume required for encryption. Then, the chaotic sequences generated by the Autapse Aihara neuron (AAN) map, combined with TSS, are used to encrypt the facial information of all authorized personnel. Finally, simulation tests and performance analyses validated the feasibility and security of the proposed encryption scheme, effectively protecting the facial information from theft and misuse.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"127 ","pages":"Article 110624"},"PeriodicalIF":4.9,"publicationDate":"2025-08-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144879239","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}