Hasil untuk "Electrical engineering. Electronics. Nuclear engineering"

Yi Jiang, Ye Bi, Yinng Li et al.

Near-eye display (NED) technology constitutes a fundamental component of head-mounted display (HMD) systems. The compact form factor required by HMDs imposes stringent constraints on optical design, often resulting in pronounced wavefront aberrations that significantly degrade visual fidelity. In addition, natural eye movements dynamically induce varying blur that further compromises image quality. To mitigate these challenges, a gaze-contingent neural network framework has been developed to compensate for aberrations within the foveal region. The network is trained in an end-to-end manner to minimize the discrepancy between the optically degraded system output and the corresponding ground truth image. A forward imaging model is employed, in which the network output is convolved with a spatially varying point spread function (PSF) to accurately simulate the degradation introduced by the optical system. To accommodate dynamic changes in gaze direction, a foveated attention-guided module is incorporated to adaptively modulate the pre-correction process, enabling localized compensation centered on the fovea. Additionally, an end-to-end trainable architecture has been designed to integrate gaze-informed blur priors. Both simulation and experimental validations confirm that the proposed method substantially reduces gaze-dependent aberrations and enhances retinal image clarity within the foveal region, while maintaining high computational efficiency. The presented framework offers a practical and scalable solution for improving visual performance in aberration-sensitive NED systems.

Lifeng Wang, Xinhua Liu, Haihua Xu et al.

Abstract Magnetorheological (MR) materials are smart materials whose rheological properties change significantly under the influence of magnetic fields. These materials mainly include fluids, elastomers, and greases. The components of MR materials consist of magnetic particles, a non-magnetic carrier liquid or matrix, and additives. The unique MR effect of these materials makes them widely used in robotics and medical devices. Improving the properties of MR materials and utilizing their characteristics are of great significance for the design and application of modern electromechanical devices. Therefore, this paper presents the composition, characteristics, and working principles of MR materials, as well as the latest progress in their applications in robotics and medical devices. Firstly, the composition and fabrication process of MR materials are introduced. Then, devices based on MR materials, including actuators, clutches, dampers, pumps, grippers for robots and medical devices, and MR robots, are extensively reviewed. Finally, a discussion of future research directions and technological challenges is provided as the conclusion of this review. The aim is to provide useful information to facilitate the design of robots and medical devices. Graphical Abstract Among the advancements of soft-bodied robots or micro-robots, MR materials exhibit the ability to deform under external magnetic fields, making them highly suitable for fabricating robots. 29 , 66 , 67 Hua et al. 68 designed an MRF-filled soft crawling robot with magnetic actuation. The robot achieves anisotropic magnetic torque-driven crawling, avoiding magnetic interference when the field is off. It demonstrates potential for applications in confined spaces, offering a novel approach to soft robotics with improved control and fabrication efficiency. McDonald et al. 69 developed an MRF-based soft robot with integrated flow control components, where motion is regulated by magnetic field-induced changes in fluid viscosity and pressure drop. This design enables multi-degree-of-freedom actuation through a single inlet and outlet, reducing the complexity of fluidic connections. The robot achieves complex behaviors such as bending, gripping, and independent control of multiple actuators, improving both autonomy and scalability while preserving compliance and adaptability. Chen et al. 70 designed a solid-liquid state transformable MRF-Robot made from an MRF. The MRF-Robot can perform diverse tasks such as large deformation, splitting, merging, object manipulation, and gradient pulling, making it suitable for biomedical applications like drug delivery and thrombus clearance. Li et al. 71 developed an untethered MRF robot encapsulated within an elastic membrane. The robot can wrap and transport delicate objects like tomatoes without causing damage, and it can also navigate through complex mazes shaped like letters. These capabilities highlight its potential for handling soft objects and operating in confined spaces. Min et al. 72 reported a stiffness-tunable, soft adhesive robot inspired by the velvet worm, utilizing an MRE that rapidly changes stiffness under an external magnetic field. This robot achieves precise adhesion control with low preload, enabling delicate grasping of soft and wrinkled surfaces without damage. The robot can unscrew nuts, and assist in mouse tumor removal surgery, showing potential in biomedical engineering (Figure 2e Figure 1: Proportional classification of MR materials applications in robots and medical devices based on Refs. [28], [29], [30], [31], [32], [33], [34], [35 Copyright (2022), Institute of Electrical and Electronics Engineers. 28 Copyright (2023), American Chemical Society. 29 Copyright (2024), Institute of Electrical and Electronics Engineers. 30 Copyright (2020), Institute of Electrical and Electronics Engineers. 31 , 34 Copyright (2022), reprinted with permission from Tan et al. 32 Copyright (2023), Institute of Electrical and Electronics Engineers. 33 Copyright (2021), Institute of Electrical and Electronics Engineers. 35 Figure(s) reproduced with permissions from the institutions/holders mentioned. ).

Fanfan Lin, Peter Wilson, Xinze Li et al.

Artificial intelligence (AI) is rapidly transforming power electronics, with AI-related publications in IEEE Power Electronics Society selected journals increasing more than fourfold from 2020 to 2025. However, the ethical dimensions of this transformation have received limited attention. This article underscores the urgent need for an ethical framework to guide responsible AI integration in power electronics, not only to prevent AI-related incidents but also to comply with legal and regulatory responsibilities. In this context, this article identifies four core pillars of AI ethics in power electronics: Security & Safety, Explainability & Transparency, Energy Sustainability, and Evolving Roles of Engineers. Each pillar is supported by practical and actionable insights to ensure that ethical principles are embedded in algorithm design, system deployment, and the preparation of an AI-ready engineering workforce. The authors advocate for power electronics engineers to lead the ethical discourse, given their deep technical understanding of both AI systems and power conversion technologies. The paper concludes by calling on the IEEE Power Electronics Society to spearhead the establishment of ethical standards, talent development initiatives, and best practices that ensure AI innovations are not only technically advanced but also oriented toward human and societal benefit.

Qi Diao, Chengyue Xie, Yuchen Yin et al.

Aiming at the shortcomings of the gazelle optimization algorithm, such as the imbalance between exploration and exploitation and the insufficient information exchange within the population, this paper proposes a multi-strategy improved gazelle optimization algorithm (MSIGOA). To address these issues, MSIGOA proposes an iteration-based updating framework that switches between exploitation and exploration according to the optimization process, which effectively enhances the balance between local exploitation and global exploration in the optimization process and improves the convergence speed. Two adaptive parameter tuning strategies improve the applicability of the algorithm and promote a smoother optimization process. The dominant population-based restart strategy enhances the algorithms ability to escape from local optima and avoid its premature convergence. These enhancements significantly improve the exploration and exploitation capabilities of MSIGOA, bringing superior convergence and efficiency in dealing with complex problems. In this paper, the parameter sensitivity, strategy effectiveness, convergence and stability of the proposed method are evaluated on two benchmark test sets including CEC2017 and CEC2022. Test results and statistical tests show that MSIGOA outperforms basic GOA and other advanced algorithms. On the CEC2017 and CEC2022 test sets, the proportion of functions where MSIGOA is not worse than GOA is 92.2% and 83.3%, respectively, and the proportion of functions where MSIGOA is not worse than other algorithms is 88.57% and 87.5%, respectively. Finally, the extensibility of MSIGAO is further verified by several engineering design optimization problems.

Florian Felten, Gabriel Apaza, Gerhard Bräunlich et al.

Engineering design optimization seeks to automatically determine the shapes, topologies, or parameters of components that maximize performance under given conditions. This process often depends on physics-based simulations, which are difficult to install, computationally expensive, and require domain-specific expertise. To mitigate these challenges, we introduce EngiBench, the first open-source library and datasets spanning diverse domains for data-driven engineering design. EngiBench provides a unified API and a curated set of benchmarks -- covering aeronautics, heat conduction, photonics, and more -- that enable fair, reproducible comparisons of optimization and machine learning algorithms, such as generative or surrogate models. We also release EngiOpt, a companion library offering a collection of such algorithms compatible with the EngiBench interface. Both libraries are modular, letting users plug in novel algorithms or problems, automate end-to-end experiment workflows, and leverage built-in utilities for visualization, dataset generation, feasibility checks, and performance analysis. We demonstrate their versatility through experiments comparing state-of-the-art techniques across multiple engineering design problems, an undertaking that was previously prohibitively time-consuming to perform. Finally, we show that these problems pose significant challenges for standard machine learning methods due to highly sensitive and constrained design manifolds.

J. Lukac, M. Kimmer, J. Sykora

The article proposes a (phase-marginalized) frame-start detector resilient to frequency offset between the transmitter and the receiver. The detector is a slight modification of the standard detector that looks for a known sequence/preamble using the matched filter, respectively correlation of the received signal with the known sequence. The modification is that the original long sequence is divided into shorter ones; the magnitude or magnitude-squared of the output of matched-filters is taken and then appropriately added to get the final detector metric. The proposed detector is a low-cost alternative to the standard approach (joint estimation of frame timing and frequency offset), where we need as many matched-filters as the number of testing frequency offsets. We derive probability characteristics of the proposed detector that reflect its resilience to the frequency offset. The frequency resilience of the detector is proportional to the number of segments into which the preamble is divided.

P. Stukhliak, O. Totosko, D. Stukhliak

The paper presents the results of a study of molecular mobility in an epoxy binder containing nanofillers of various types: nanodispersed copper (Cu), cobalt (Co), and carbon multilayer nanotubes (CMLN) under modification by electric spark hydro-impact (ESHI). The research focuses on the dielectric characteristics of the formed composites, in particular on the dielectric loss tangent, which serves as an informative indicator of segmental mobility of polymer chains and the intensity of relaxation processes in the solid state. It has been established that the introduction of nanofillers into the epoxy matrix leads to a systematic shift of the temperature maximum of the dielectric loss tangent and to a change in the shape of relaxation peaks, which indicates a modification of the molecular mobility of macromolecular fragments and the formation of interphase layers with altered energy states. Comparative analysis of the influence of Cu, Co, and CMLN reveals different efficiencies in the formation of interfacial interactions, determining the nature of dipole relaxation, the degree of restriction of segmental motion, and the polarization behavior of the composite. Carbon multilayer nanotubes demonstrate the most pronounced effect on relaxation processes due to the formation of a branched conductive network and barrier layers, whereas metallic nanoparticles mainly influence local polarization mechanisms associated with surface defects and electronic polarization. The obtained results confirm that ESHI modification combined with nanofilling enables controlled tuning of dielectric and relaxation properties of epoxy systems. This creates prerequisites for the development of functional polymer composites with predetermined electrical, polarization, and operational characteristics for applications in electrical engineering, electronics, and protective coatings.

R. of, G. E. A. E. E. 1, W. Electronics et al.

In this paper, unbalanced three-phase fault in transmission lines is considered with respect to estimating the state of power system after a fault occurs at different buses. Faults such as a single-line-to-ground (SLG), line-to-line (LL) and double-line-to-ground (DLG) affect the bus system that is connected along with the transmission line. MATLAB software was employed in which unbalanced fault programs based on the Symmetrical Component method to determine the voltage magnitudes, line current magnitude, total fault current, real and reactive power at Phase A, Phase B and also on phase C for the different bus lines. The unbalanced fault programs are executed using a Newton Raphson based power flow program for the standard IEEE 14, IEEE 26 and IEEE 30 bus systems. The obtained results show that the single line to ground fault is the most severe kind for IEEE 14 bus system, while for IEEE 26 and IEEE 30 bus system, the most severe fault is line to line fault. This finding is crucial for evaluating the reliability and stability of power transmission lines.

Yao Fu, Rameswar Panda, Xinyao Niu et al.

We study the continual pretraining recipe for scaling language models' context lengths to 128K, with a focus on data engineering. We hypothesize that long context modeling, in particular \textit{the ability to utilize information at arbitrary input locations}, is a capability that is mostly already acquired through large-scale pretraining, and that this capability can be readily extended to contexts substantially longer than seen during training~(e.g., 4K to 128K) through lightweight continual pretraining on appropriate data mixture. We investigate the \textit{quantity} and \textit{quality} of the data for continual pretraining: (1) for quantity, we show that 500 million to 5 billion tokens are enough to enable the model to retrieve information anywhere within the 128K context; (2) for quality, our results equally emphasize \textit{domain balance} and \textit{length upsampling}. Concretely, we find that naively upsampling longer data on certain domains like books, a common practice of existing work, gives suboptimal performance, and that a balanced domain mixture is important. We demonstrate that continual pretraining of the full model on 1B-5B tokens of such data is an effective and affordable strategy for scaling the context length of language models to 128K. Our recipe outperforms strong open-source long-context models and closes the gap to frontier models like GPT-4 128K.

Satyabrat Rath, Jothi Ramalingam

The outsourcing protocols for Eigen-Decomposition (ED) and Singular Value Decomposition (SVD) proposed by Zhou and Li (2016) offer intriguing advancements but are susceptible to malicious behavior by cloud entities. Our investigation identifies a critical vulnerability in the verification scheme utilized by Zhou and Li, where a malicious cloud can deceive the client by providing incorrect results that pass the verification step undetected. This paper not only demonstrates this vulnerability through a detailed attack scenario but also proposes an enhanced verification method to fortify the protocols against such malicious activities, ensuring the integrity and reliability of the schemes proposed by Zhou and Li.

Ali Mansoor, X. Diao, C. Smidts

Abstract The increased complexity of modern system designs and demands for quicker time to market have made safety-related verification and validation of such systems more challenging. Incorporating safety and risk considerations at the early stages of design is one way to acquire a more robust initial design for novel systems. Inductive fault analysis has its significance at final stages of design, e.g., verification and validation. However, to preclude certain system failure states—especially for the systems with high failure consequences, a designer would innately prefer to trace back and remedy the causes of failure, as compared to a more cumbersome activity of identifying the faults individually and sifting the combinations that lead to the failure of interest. The work presented in this paper is aimed at the development of a backward failure propagation methodology for analyzing the origins of functional failures in a conceptual design of systems including but not limited to nuclear, mechanical, aerospace, process, electrical/electronics, telecommunication, automotive, etc. This method allows the designer to achieve a robust early design based on the analyses of the system’s functional dependencies before proceeding to the detailed design and testing stages. The insights provided by the analysis at the conceptual design stage also reduce redesign efforts, testing costs, and project delays. The proposed method is a functional analysis approach that extends the Integrated System Failure Analysis for backward failure propagation. When provided with an abstract system configuration, a system’s functional model, and a system’s behavioral model, it utilizes a known functional state (typically a failure) to acquire system component modes and the states of other functions. The method includes inversion of the functional failure logic and component behavioral rules using propositional logic and deductive analysis to assess valid states of a system that satisfy the given initial conditions. To test the method’s scalability, we applied the proposed method to a simplified representation of the secondary loop of a typical pressurized water reactor.

D. S. Parno, A. W. P. Poon, V. Singh

There are profound connections between neutrino physics and nuclear experiments. Exceptionally precise measurements of single and double beta-decay spectra illuminate the scale and nature of neutrino mass and may finally answer the question of whether neutrinos are their own antimatter counterparts. Neutrino-nucleus scattering underpins oscillation experiments and probes nuclear structure, neutrinos offer a rare vantage point into collapsing stars and nuclear fission reactors, and techniques pioneered in neutrino nuclear-physics experiments are advancing quantum-sensing technologies. In this article, we review current and planned efforts at the intersection of neutrino and nuclear experiments.

Lisi Zhang, Qiancheng Yu, Beijing Zhou et al.

The efficacy of credit risk assessment models is pivotal to the risk management capacity of financial institutions. Traditional credit risk models often suffer from inadequate predictive accuracy due to overlooked feature combinations and weak supervisory signals. Addressing these limitations, we present a novel approach for credit default prediction that integrates feature interactions and contrastive learning. Specifically, we introduce second-order interactions atop standard linear models to achieve low-order feature interplay. Concurrently, the integration of deep neural networks and attention mechanisms facilitates the learning of concealed high-order features, thus enhancing the model’s non-linear modeling capabilities and illuminating latent feature associations. Further, to ameliorate the issues of noise and diminished supervisory signals, we embed slight noise in feature embeddings for data augmentation and construct contrastive views, ultimately refining feature quality. To attest to the effectiveness of our approach, we conducted experiments on two real-world datasets, benchmarking against eight predictive methods including LR, XGBoost, and FiBiNET. The results unequivocally demonstrate the superior performance of our method across various metrics, underscoring its promise and excellence in the realm of credit risk assessment.

Sumit Bhardwaj, Aayush Gupta, G. Kumar et al.

Wen-chuan Wang, Lei Xu, K. Chau et al.

This work is directed towards the improvement of programming skills for students on the 2nd semester of Electrical Engineering (EE) BEng signed up for the “Digital Electronics and Programming” (DEP) course. The goal is to support students who tend to drop out in the first half of the semester or give up and not show up at oral exams. In our research hypothesis we state: Decreasing and eliminating negative emotional experiences will increase student’s self-efficacy thereby lowering dropout. We experiment with a programming process guide in text and video to gain students' capability in solving programming problems and increase metacognitive awareness. Additionally, we measure students' emotional experience of programming by using a special self-assessment vignette inquiry. For comparison purposes we introduced the same measurement methodology in 2021 on the courses where there is no focus on the process guideline. The results show a positive effect on lowering the negative emotional impact when students use a programming process guide. The article describes the theoretical background based on literature studies for both the process and the students’ self-assessment, and discusses results achieved in relation to the findings in the literature. The preliminary results are described in [8] and in paper here we present recent outcomes for the autumn 2021 and spring 2022 semester over 9 weeks.

Juhyeon Kim, Emin Orhan, Kijung Yoon et al.

Neurons in the brain are complex machines with distinct functional compartments that interact nonlinearly. In contrast, neurons in artificial neural networks abstract away this complexity, typically down to a scalar activation function of a weighted sum of inputs. Here we emulate more biologically realistic neurons by learning canonical activation functions with two input arguments, analogous to basal and apical dendrites. We use a network-in-network architecture where each neuron is modeled as a multilayer perceptron with two inputs and a single output. This inner perceptron is shared by all units in the outer network. Remarkably, the resultant nonlinearities often produce soft XOR functions, consistent with recent experimental observations about interactions between inputs in human cortical neurons. When hyperparameters are optimized, networks with these nonlinearities learn faster and perform better than conventional ReLU nonlinearities with matched parameter counts, and they are more robust to natural and adversarial perturbations.

肖丙政, 谢海平, 魏刚武 et al.

The production process of steelmarking bearing steel by 100 t converter is studied and analyzed with tlhermody namics. The co-operating control process of carbon preservation, dephosphorization, and temperature control (C-P-T) in converter steelmaking process is optimized and established and applied to produce bearing steel. The results show that the dephosphorization reaction in furnace does not stop immediately when the selective oxidation temperature of phosphorus and carbon comes but has a slow decay process. In order to meet the need of deep dephosphorization, the optimal time of first slagging out in early stage of the new process should be controlled between 350 s and 380 s, and the temperature range is 1 360-1 437 ℃. The decarburization rate model predicts that the decarburization rate of bearing steel in the late stage of smelting is (0. 21% -0. 28% )/min. In order to keep the requirement of carbon preservation and temperature control, the catch carbon and reblow can be adopted in the end phase of blowing to further accurately control the temperature of liquid steel. According to the production data statistics of a furnace service period, the heat ratio of end carbon, phosphorus and temperature hit the target at the same time accounts for 76.67% , and the purity of liquid steel is greatly improved.

Xuemei Fu, Haitao Yang, Zhipeng Li et al.

Emerging wearable electronics, wireless communication, and tissue engineering require the development of conductive fiber-shaped electrodes and biointerfaces. Ti3C2Tx MXene nanosheets serve as promising building block units for the construction of highly conductive fibers with integrated functionalities, yet a facile and scalable fabrication scheme is highly required. Herein, a cation-induced assembly process is developed for the scalable fabrication of conductive fibers with MXene sheaths and alginate cores (abbreviated as MXene@A). The fabrication scheme of MXene@A fibers includes the fast extrusion of alginate fibers followed by electrostatic assembly of MXene nanosheets, enabling high-speed fiber production. When multiple fabrication parameters are optimized, the MXene@A fibers exhibit a superior electrical conductivity of 1083 S cm-1, which can be integrated as Joule heaters into textiles for wearable thermal management. By triggering reversible de/hydration of alginate cores upon heating, the MXene@A fibers can be repeatedly contracted and generate large contraction stress that is >40 times higher than the ones of mammalian skeletal muscle. Furthermore, the MXene@A springs demonstrate large contraction strains up to 65.5% and are then fabricated into a reconfigurable dipole antenna to wirelessly monitor the surrounding heat sources. In the end, with the biocompatibility of MXene nanosheets, the MXene@A fibers enable the guidance of neural stem/progenitor cells differentiation and the promotion of neurite outgrowth. With a cation-induced assembly process, our multifunctional MXene@A fibers exhibit high scalability for future manufacturing and hold the prospect to inspire other applications.

Armin Moin

This doctoral dissertation proposes a novel approach to enhance the development of smart services for the Internet of Things (IoT) and smart Cyber-Physical Systems (CPS). The proposed approach offers abstraction and automation to the software engineering processes, as well as the Data Analytics (DA) and Machine Learning (ML) practices. This is realized in an integrated and seamless manner. We implement and validate the proposed approach by extending an open source modeling tool, called ThingML. ThingML is a domain-specific language and modeling tool with code generation for the IoT/CPS domain. Neither ThingML nor any other IoT/CPS modeling tool supports DA/ML at the modeling level. Therefore, as the primary contribution of the doctoral dissertation, we add the necessary syntax and semantics concerning DA/ML methods and techniques to the modeling language of ThingML. Moreover, we support the APIs of several ML libraries and frameworks for the automated generation of the source code of the target software in Python and Java. Our approach enables platform-independent, as well as platform-specific models. Further, we assist in carrying out semiautomated DA/ML tasks by offering Automated ML (AutoML), in the background (in expert mode), and through model-checking constraints and hints at design-time. Finally, we consider three use case scenarios from the domains of network security, smart energy systems and energy exchange markets.