Search Results (1,790)

Effective evaluation of building fabric is essential to understanding building performance. Traditional methods, such as ISO 9869-1, provide point U-values but do not always capture the complete picture, and new measurement-led approaches are needed to support building retrofit. Heat3D is a novel iOS application that performs rapid U-value measurements of building elements by mapping thermographic images from a mobile infrared camera onto an augmented reality (AR) model of a room, from which heat flux across the element is calculated. A field trial of 22 UK properties during the 2019–2020 winter heating season assessed Heat3D’s heat flux measurements against traditional heat flux plates, with 90% of 295 Heat3D surveys falling within the combined uncertainty of the reference method. A second field trial, conducted over the 2020–2021 winter heating period, evaluated a new timelapse iteration of the method capable of measuring wall U-value within approximately 60 min; 90% of 42 Heat3D surveys fell within the combined confidence interval of the ISO 9869-1-measured U-value. These results indicate that Heat3D offers a viable, rapid alternative to traditional methods for both heat flux and U-value measurement. Full article

Large language models (LLMs) are increasingly applied across drug discovery, yet their role in analytical method development to support quality control in cell-based potency assays remains insufficiently explored. This study evaluates whether general-purpose LLMs can assess plate homogeneity and detect spatial bias using a half-maximal effective concentration (EC₅₀) mapping approach implemented in a 96-well cell-based reporter gene assay. Four independent plates were generated by two analysts on different days and analyzed using a conventional spreadsheet workflow and three LLMs (Google Gemini, ChatGPT, and Microsoft Copilot Analyst) under identical prompt conditions. All approaches consistently identified a statistically significant row-wise positional effect with no significant column-wise effects, supported by one-way analysis of variance (ANOVA), confidence-interval summaries, and heat-map visualizations of raw signal and Z-scores. The dominant top-to-bottom signal gradient was reproducible across individual plates and the averaged dataset. While all LLMs reproduced standard plate quality control (QC) metrics, they differed primarily in workflow completeness, responsiveness to iterative prompting, and in the extent to which additional spatial diagnostics were proposed. Overall, these results suggest that general-purpose LLMs can reproduce conventional plate-effect analyses and may serve as practical analytical companions when guided by structured prompts and complete datasets. Full article

UAV inspection images of transmission line spacers often contain small-scale defect targets, complex backgrounds, and weak fine-grained features. We propose MFIA-YOLO, a small-object defect detection model for spacer defects. Using the lightweight YOLOv8n variant as the baseline, we introduce a Multi-scale Spatial Heterogeneous Convolution (MSHC) into the backbone network. This module enhances the extraction of multi-scale features from spacer defect targets. Before the SPPF module, we further construct a Feature Complementary module (FCM). The FCM embeds shallow spatial location information into deep semantic features, thereby alleviating spatial information degradation during small-object detection. In the detection head, a C2f_IAFF module is adopted to adaptively fuse features at different scales through iterative attention-based feature fusion. This design improves the representation of defect targets in complex scenes. In addition, a transmission line spacer defect dataset is constructed from UAV inspection images collected in Xilingol League, Inner Mongolia. Experimental results showed that MFIA-YOLO achieved an mAP50 of 96.53% and an mAP50–95 of 83.50%. Compared with representative YOLO-series models, MFIA-YOLO achieved a better balance between detection accuracy and model complexity. These results demonstrate its effectiveness for accurate spacer defect detection in transmission lines. Full article

We establish a metrization result for some $b_v\left(s\right)$-metric spaces, extending a result recently established for rectangular b-metric spaces. Furthermore, we introduce the notion of an extended polygonal b-metric space (or $b_v\left(\theta \right)$-metric space), which unifies and generalizes several classes of spaces, including metric spaces, rectangular metric spaces, b-metric spaces, rectangular b-metric spaces, polygonal metric spaces, and $b_v\left(s\right)$-metric spaces. Some fixed-point results in $b_v\left(\theta \right)$-metric spaces are established under the weak orbital completeness condition in the framework of the Banach contraction principle and for generalized expansive Hardy-Rogers-type mappings. An a priori error estimate for the iterative process is obtained in both $b_v\left(\theta \right)$-metric and $b_v\left(s\right)$-metric spaces. We also establish the Ulam-Hyers stability of fixed-point equations in both $b_v\left(\theta \right)$-metric and $b_v\left(s\right)$-metric spaces. Several examples are provided, and applications to certain types of integral equations and initial value problems are presented, illustrating the applicability and effectiveness of the obtained results. Full article

Distributed acoustic sensing (DAS) is increasingly used for third-party intrusion (TPI) detection in oil and gas pipeline monitoring, but labeled DAS data are often scarce, leading to overfitting, poor generalization, and increased false alarms and missed detections. Conventional data augmentation, GAN-based synthesis, and transfer learning may generate physically implausible samples or fail to cover the event feature space. To address this, we propose a physics-guided large language model (LLM) prompt-engineering framework for DAS data augmentation and pipeline intrusion detection. The framework establishes a physically grounded feature-indicator framework for DAS disturbance-event classification by mapping primary event mechanisms to measurable signal indicators, and then uses a standardized four-module prompt template to guide LLM-based synthesis-script generation. A two-stage iterative verification procedure is further introduced to constrain the generated samples in terms of physical-mechanism compliance and feature-parameter consistency. Synthetic data are combined with real data to train a lightweight PatchTransformer model for TPI detection, while an additional CNN is used to assess cross-architecture applicability. Using the public DAS1K benchmark with five-fold stratified cross-validation and a univariate controlled experiment (0–800 synthetic samples per category), the results show that the use of synthetic data improves detection performance overall. The configuration with 600 synthetic samples per category achieves 92.27% accuracy and 92.38% macro-F1, outperforming the conventional augmentation baseline by 4.74 and 4.86 percentage points, respectively. An additional CNN experiment also showed consistent performance gains across the tested augmentation settings, indicating that the benefit of the proposed synthetic data was not restricted to the PatchTransformer architecture. These findings indicate that LLM-assisted data augmentation can effectively improve the generalization of DAS-based pipeline intrusion detection when field-labeled samples are scarce. Full article

Dysgeusia is a clinically consequential, but still under-standardized, toxicity of cancer treatment. In the immunotherapy era, taste disturbances are increasingly relevant for patients receiving immune checkpoint inhibitors, chimeric antigen receptor (CAR) T-cell therapies and T-cell-redirecting bispecific antibodies, with G protein-coupled receptor family C group 5 member D (GPRC5D)-directed treatment in multiple myeloma representing a particularly instructive high-burden model. We performed a structured critical narrative review with evidence mapping. PubMed/MEDLINE was searched from database inception to June 2026, complemented by citation tracking in Google Scholar, ClinicalTrials.gov searches and guideline documents relevant to oncology nutrition, oral supportive care and cancer-related taste dysfunction. Search concepts covered cancer-related dysgeusia, immunotherapy-associated oral toxicity, GPRC5D/talquetamab-associated dysgeusia, oncology nutrition, oral–gut microbiome biology, natural taste modulators and miraculin-based interventions. Dysgeusia can reduce appetite, food enjoyment, dietary diversity and protein energy intake, thereby contributing to weight loss, malnutrition risk, distress, social withdrawal and, in severe cases, treatment modification or discontinuation. Available evidence is heterogeneous: general cancer-treatment-associated dysgeusia is supported by broader observational and interventional literature; immunotherapy-associated dysgeusia is less systematically characterized; and GPRC5D/talquetamab-associated dysgeusia represents the most clinically visible and target-specific immunotherapy-associated phenotype. Emerging pilot data suggest that dried miracle berry or miraculin-containing products may improve selected taste perception and nutritional parameters in cancer-related dysgeusia, but direct evidence in immunotherapy-associated dysgeusia is not yet established. We, therefore, propose a claim-disciplined precision supportive-care framework integrating systematic taste phenotyping, early nutritional risk assessment, oral health evaluation, microbiome-aware but hypothesis-generating endpoints, individualized flavor and texture adaptation, cautious use of natural taste modulators in selected patients and iterative monitoring of patient-centered outcomes. Future trials should test whether dysgeusia-focused nutritional and taste-modulating supportive care interventions can improve intake, quality of life and treatment persistence without compromising immunotherapy safety or efficacy. Full article

Accurate detection of loess landslide hazards is critical for disaster prevention, yet remains challenging due to complex terrain, spectral similarity to background regions, and the lack of specialized datasets. To address these challenges, this paper proposes LHEM-MSegNet, a multisource landslide hazard extraction model based on Transformer and attention fusion. A loess landslide hazard region segmentation dataset for Linyou County (LYPLH), Shaanxi Province, China, is constructed using a transfer-learning strategy based on the Segment Anything Model (SAM), followed by iterative manual correction and data augmentation to improve annotation quality. LHEM-MSegNet integrates Transformer blocks with spatial and channel attention modules to enhance feature representation, and adopts a multi-pathway architecture for multisource fusion. Specifically, optical remote sensing imagery is combined with geomorphological information, including DEM, slope, and aspect maps, to improve landslide hazard extraction. Experimental results on the LYPLH dataset demonstrate that the proposed LHEM-MSegNet achieves an mIoU of 89.29% by integrating optical imagery with geomorphological information. Ablation studies first demonstrate the effectiveness of the proposed model structure and further confirm the significant contribution of terrain-related features to landslide hazard extraction. In addition, the supplementary evaluation on the external Landslide4Sense benchmark shows that the proposed method maintains competitive performance under different data distributions. The proposed approach provides an effective solution for early loess landslide hazard identification and supports disaster mitigation in the Loess Plateau region. Full article

Parameter optimization of injection and production is an important method used to enhance recovery rate and reduce water cut, mainly aiming to determine the injection and production strategies during the development process to maximize the economic benefits throughout the reservoir development process. However, current optimization approaches for injection and production face challenges such as complex and inefficient optimization models and high-dimensional discrete variables, making it difficult to improve the global optimization ability of the algorithm (avoiding local optimal solutions during the optimization process) and the controllability of the time for completing the optimization of injection and production parameters under real and limited numerical simulations. This paper proposes a high-dimensional multi-discrete injection and production parameter non-gradient optimization method (MNOM), which combines the upper confidence bound (UCB) algorithm and effectively explores better injection and production systems and small-layer water-allocation schemes, achieving the maximization of net present value (NPV) over the entire development period. Specifically, this method models the injection and production parameter optimization problem as a Monte Carlo tree search process (Monte Carlo tree search, MCTS), and implements the optimization of injection and production cycles and small-layer water allocation through a genetic algorithm (CLGA) proxy optimized by a convolutional long short-term memory network (ConvLSTM). This method effectively overcomes the spatial and temporal limitations of the search process, maps production dynamics to the random strategies of injection and production parameters, and estimates the expected return of each policy. The CLGA proxy rapidly identifies suitable well-control schedules and water-allocation schemes in real time based on the production status at different development stages, thereby improving overall production performance. The proposed method has two innovative points. Firstly, MCTS can explore the large-scale discrete space of injection and production parameter optimization variables through tree decomposition, combined with the UCB incentive mechanism, to improve the global optimization ability. Secondly, the model training process is completely based on existing physical laws, with good temporal evolution, and the trained strategy can quickly adapt to the production status of the target layer without the need for a complete re-training from the beginning, enabling offline application and having good real-time controllability. In order to verify the effectiveness of the method proposed in the article, tests were conducted on a 3D reservoir actual model. Compared with gradient-based approaches, classical evolutionary algorithms, and proximal policy optimization (PPO), MNOM not only achieves stronger global search performance and requires 55–78% fewer iterations, but also improves the effective sweep volume by 1.1% to 5.5% compared to other optimization methods; furthermore, when compared with the PPO method, it is found that in offline optimization, if the production regime changes, the training strategy has better real-time controllability. The MNOM method can still maintain the original optimization effect compared to the PPO method when the production regime changes, demonstrating better engineering adaptability. The research results show that the proposed multi-strategy fusion high-dimensional multi-discrete injection and production parameter non-gradient optimization method can effectively improve recovery rate, expand effective sweep volume, and balance global optimization ability, optimization efficiency, and real-time controllability under the constraint of limited numerical simulations, and it has good engineering adaptability and application prospects. Full article

Generating synthetic aperture radar (SAR) images at unknown azimuth angles under limited sample conditions remains a challenging task, since target scattering characteristics vary significantly with observation angle and are difficult to model effectively using only image-domain priors. To address this issue, this paper proposes a scattering center prior-guided conditional diffusion framework for unknown-azimuth SAR image generation. First, the Iterative Shrinkage–Thresholding Algorithm (ISTA) constrained by the Point Spread Function (PSF) is employed to extract dominant scattering centers from SAR images at known viewing angles, obtaining sparse and physically interpretable scattering center maps. Subsequently, the target category, azimuth angle, and scattering map are jointly used as conditional vector inputs to train the diffusion model. During the inference stage, scattering maps from known azimuth angles are fused to construct a scattering prior for the unknown azimuth, which is used to guide the generation of the corresponding SAR image. Experimental results under sparse angular sampling conditions demonstrate that, compared with the scattering-guided GAN baseline and the non-learning image-domain interpolation baseline, the proposed method better preserves dominant scattering structures and generates unknown-azimuth SAR images with clearer target contours, more stable strong scattering regions, and fewer local artifacts. In summary, introducing dominant scattering center priors into the conditional diffusion model provides effective physical constraints for SAR image generation and improves unseen-azimuth SAR image generation under the evaluated sparse angular sampling conditions. Full article

Traditional motion compensation (MoCo) methods for high-resolution synthetic aperture radar (SAR) rely on iterative processing between imaging and MoCo, incurring heavy computational overhead and barely meeting fast imaging requirements. To address this, this paper presents a non-iterative SAR autofocus method based on multi-subimage two-dimensional phase gradient autofocus (2D-PGA). Leveraging the parallel imaging capability of the ground Cartesian back-projection (GCBP) algorithm, we reveal an a priori 2D spatially variant spectral structure of GCBP subimage errors, then the proposed 2D-PGA method is utilized to precisely estimate subimage wavenumber-domain errors. Mapping between subimage offsets and linear phase errors is derived for cross-subimage error splicing. The spliced errors are compensated for corresponding subimages to realize autofocus and ensure coherent subimage fusion, enabling non-iterative parallel processing of SAR time-domain imaging and MoCo. Experiments with 0.03 m-resolution Ku-band microwave photonic SAR measured data verify the necessity and effectiveness of the proposed method. Full article

We formulate and analyze a 26-dimensional nonlinear dynamical system governing a doubly-fed induction generator (DFIG) wind energy conversion system coupled to an infinite bus through a dynamic transmission line. Seven interacting subsystems—aerodynamics, a two-mass drivetrain, a fourth-order machine, rotor- and grid-side converter controllers, a phase-locked loop, and a pitch regulator—are assembled into a single vector field $\dot{\mathbf{x}}=\mathbf{f}(\mathbf{x},\mathbf{u})$ on ${\mathbb{R}}^{26}$, derived in dimensionless coordinates. Strict positivity of the determinant $\mathrm{\Delta }=L_s{L}_r-L_m^2=\sigma L_s{L}_r$ for every physically admissible machine renders the flux–current map invertible, so the right-hand side is well defined; the nodal Kirchhoff constraint forms a semi-explicit differential-algebraic relation that we eliminate to obtain an explicit ordinary differential equation. The central contribution is a constructive scheme for the equilibria: the 26 stationarity conditions $\mathbf{f}({\mathbf{x}}^{\star },\mathbf{u})=\mathbf{0}$ are solved by an iterative voltage-matching procedure converging to a residual below ${10}^{-11}$ per unit—essentially machine precision—which removes the spurious start-up transients common in reported simulations. Analytically chosen feedback gains induce a hierarchy of well-separated time scales, placing the closed loop in the multiple-time-scale class; the separation is made quantitative through explicit small parameters ${\epsilon }_i$ formed from the ratios of subsystem time constants. Numerical integration of a GE $3.6$ MW configuration confirms the construction: under stationary forcing, the rotor speed stays within $1.32\times {10}^{-5}$ pu of the equilibrium, and under a large-amplitude wind program ($11\to 14\to 9$ m/s) spanning the full operating envelope, it is regulated to within $0.065\%$, while the DC-link voltage deviation remains below $2.4\times {10}^{-5}$ pu and the power balance closes with residual below ${10}^{-3}$ pu, the ≈2% mechanical–electrical gap being the modeled losses. Linearization about the computed equilibrium yields a Jacobian whose spectrum lies entirely in the open left half-plane, establishing local asymptotic stability and exposing the individual electromagnetic, torsional, and control modes. The model furnishes a rigorously initialized, analytically transparent basis for linearization, spectral stability analysis, and bifurcation study. Its practical value is that a consistent equilibrium and a certified spectrum remove the start-up transients and undocumented tuning that otherwise let initialization artifacts masquerade as genuine dynamics, so that the model can serve as a trustworthy building block for weak-grid and wind-farm stability studies. Full article

Flexible metamaterial absorbers have exhibited tremendous potential for applications in intelligent wearable devices and radar stealth protection due to their remarkable electromagnetic response characteristics and mechanical conformal adaptability. However, conventional metamaterial development relies heavily on iterative full-wave simulations, which not only incurs prohibitive computational costs but also hinders the efficient identification of global optima within high-dimensional geometric parameter spaces. To address these challenges, this paper proposes an inverse design framework based on a deep learning-powered adversarial diffusion model. By integrating residual blocks and self-attention mechanisms within the U-Net architecture, the model’s capacity to capture global spectral features is significantly enhanced. Furthermore, the introduction of a discriminator for adversarial fine-tuning optimizes generation quality, resulting in a 22.46% reduction in the target loss function compared with conventional approaches. This method effectively resolves the “one-to-many” inverse mapping challenge between spectral requirements and geometric structures. Experimental results demonstrate that the designed absorber exhibits excellent polarization insensitivity and maintains efficient, stable absorption performance even under large-angle conformal bending. Moreover, a multi-sample collaborative validation strategy is employed to cross-verify measured samples across different frequency bands, establishing the model’s high precision and engineering reliability. Full article

This paper introduces the Contraction Mapping Optimizer (CMO), a population-based optimizer whose update rule is derived directly from the Banach fixed-point theorem rather than from a biological or physical metaphor, and for which global convergence is proved. Most metaheuristics are built on such metaphors and offer little formal insight into why their update rules converge; CMO instead makes the mechanism explicit. Each candidate solution is moved by a damped contraction map toward an attractor formed from the best solutions found so far, perturbed by a Gaussian exploration term whose amplitude vanishes over the run. Because the contraction factor is kept strictly below one by construction, the deterministic part of every update is provably a contraction, making the exploration-exploitation balance explicit and schedulable. We further prove that the full stochastic, population-based iteration converges to the global optimum, and we demonstrate CMO on a nonlinear integral equation whose solution is itself a fixed point. CMO is compared with nine established metaheuristics (GA, PSO, DE, GWO, HHO, RUN, EO, FGO and WOA) on CEC2017 (D = 30, 50, 100) and CEC2022 (D = 10, 20), on three constrained engineering design problems, and on a six-unit economic load dispatch problem, over thirty runs at a common budget. Pooled over the 111 benchmark instances, CMO attains a Friedman mean rank of 3.14, statistically inseparable from the best competitor (FGO, 2.20) and significantly ahead of the remaining seven baselines; it ranks first on both CEC2022 suites and, on the constrained applications, is second only to differential evolution. Sensitivity and ablation studies identify the vanishing exploration schedule and greedy selection as the decisive components of the design. Full article

This paper investigates a class of nonlinear fractional differential equations with boundary value problems involving Caputo–Hadamard-type derivatives. By relaxing the monotonicity constraints on the nonlinear terms and considering more general nonlinear structures, the paper extends the theoretical framework and application scope of the relevant fractional models. Using the upper-lower solution method in conjunction with Schauder’s fixed-point theorem, we establish the existence of exact solutions; furthermore, by applying the Banach contraction mapping principle, we prove the uniqueness of the solutions. Concurrently, we construct a convergent iterative approximation scheme and provide a priori and a posteriori error estimates for numerical solution. Furthermore, the robustness of the solutions to perturbations is characterised via Ulam–Hyers stability analysis, ensuring the reliability of the approximate solutions. Finally, numerical examples are employed to validate all theoretical results. The qualitative theory and numerical analysis methods for Caputo–Hadamard-type fractional boundary value problems have been improved. Full article

This study develops the Socio-Technical Impact (ST-Impact) model, a sustainability-oriented curriculum design framework for management information systems (MIS) education, aimed at preparing future professionals to design and govern digital systems in socially and environmentally responsible ways. The rapid diffusion of emerging technologies, particularly generative AI, has intensified the need for MIS professionals who can integrate technical competence with social responsibility, ethical reasoning, responsible digital design, and public-value-oriented problem solving—competencies that remain unevenly addressed in existing MIS curricula. To address this gap, the study adopts a design science research approach, conducting a systematic review, reported in accordance with PRISMA 2020, and a design-oriented narrative synthesis of 120 international studies published between 2010 and 2025. Learning activities, student-produced artifacts, and assessment mechanisms were extracted as curriculum design units, while governance-related indicators were coded according to stakeholder requirements, accountability, equity, accessibility, privacy, safety, explainability, and sustainability. The resulting ST-Impact model comprises six iterative modules and a three-layer evaluation system. Its design coherence and practical plausibility are examined through an evidence-to-model traceability mapping, an illustrative comparative analysis of publicly visible curriculum structures, and an adaptable 15-week syllabus architecture. By translating abstract concepts of societal impact, responsible digital design, and digital sustainability into actionable curriculum design elements, this study contributes a literature-grounded foundation for future empirical validation in MIS education. Full article