Search Results (1,871)

The reconstruction of nodal prices and generation patterns in electricity markets with network constraints constitutes a challenging inverse analysis problem due to congestion-induced non-uniqueness and limited observability. This study introduces node symmetry analysis as a novel early indicator of locational marginal price (LMP) growth in power systems with high renewable energy penetration. Symmetric nodes, defined as nodes with identical generation cost structures and comparable network topology, exhibit near-identical price signals under uncongested conditions. In this study, the term “price” refers to the LMP obtained from the DC-OPF market-clearing model under scenarios with high renewable energy penetration. Deviations from this symmetry, quantified through price differences between symmetric node pairs (ΔLMP), serve as sensitive indicators of emerging network stress and congestion, providing early warning of peak-price events. Using DC power flow sensitivities and congestion indicators, LMPs are reconstructed in a simplified five-node test system under three scenarios: baseline operation, severe transmission congestion, and high renewable generation variability. Results show strong correlations between symmetry violations and system-wide price increases. In congested scenarios, ΔLMP exceeding €2/MWh consistently precedes peak prices by 1–2 h, demonstrating the metric’s predictive capability. Integration of storage further highlights the operational value of symmetry-based analysis, showing reductions in curtailed renewable generation and peak prices. The proposed framework offers a computationally efficient and interpretable tool for congestion diagnosis, price trend forecasting, and inverse market analysis, with potential scalability to larger AC networks and stochastic scenarios. These findings provide actionable insights for system operators, market participants, and regulators seeking to enhance flexibility, reliability, and economic efficiency in high-renewable electricity markets. Full article

To address the challenge of coarse aggregates hindering steel fiber dispersion and reducing toughening efficiency in ultra-high-performance concrete containing coarse aggregate (UHPC-CA), this study proposes a hybrid fiber design method based on reverse adaptation to the aggregate structure: a paradigm where fiber proportions are inversely designed to match the quantified void size distribution within the coarse aggregate skeleton. Industrial X-ray computed tomography (X-CT) was employed to capture the internal structure of UHPC-CA. Digital image processing techniques were used to quantitatively characterize the size distribution within the coarse aggregate skeleton gap. Based on this distribution, the blending proportions of multi-scale (3–16 mm) copper-plated steel fibers were systematically determined. Three fiber configurations were compared: mono-sized 13 mm fibers (Type A), an empirical model based on aggregate size (Type B), and a quantitatively designed blend based on skeleton gap distribution (Type C). At the same fiber volume fraction, the mechanical property test results show that the C type achieves approximately 18.6% higher flexural strength and 29.1% higher splitting tensile strength compared to the A type, while showing 5.3% and 6.7% improvements over the B type, and the compressive strength also increased slightly (about 3.0%). The microanalysis further confirms that the fiber distribution in the C-type design was more uniform, and the bridging effect and crack resistance were more sufficient. The proposed gap-adaptive fiber design paradigm offers an effective approach for optimizing reinforcement distribution in composites, providing theoretical and practical value for high-performance UHPC-CA applications. Full article

Forest height is a key structural parameter for characterizing forest architecture and estimating carbon storage. However, under complex terrain and heterogeneous forest conditions, Polarimetric synthetic aperture radar (PolSAR)-based forest height inversion using multi-category features still faces several challenges, including feature redundancy, insufficient characterization of the nonlinear couplings among high-dimensional features by deep learning models, and the difficulty of jointly achieving model stability and interpretability. In this paper, to address these issues, we propose a method for SHapley Additive exPlanations (SHAP) interpretability-driven PolSAR forest height inversion based on deep learning and multi-category feature fusion. Firstly, a deep neural network (DNN) is constructed, and SHAP is introduced to interpret the model decision process, enabling the identification of key feature interactions with clear physical significance and guiding the iterative model optimization in an explainability-driven manner. Furthermore, a SHAP-guided feature attention DNN is developed, in which the feature contribution scores are incorporated as prior knowledge for attention weight initialization, thereby establishing a closed-loop modeling framework from “interpretation” to “optimization”. Experiments were conducted at the site of the Huangfengqiao forest farm, Youxian County, Hunan province, China, using ALOS-2 L-band fully polarimetric SAR imagery. The experimental results demonstrated that the proposed method can significantly outperform the conventional machine learning approaches and various deep learning architectures for forest height inversion. The final model achieved a coefficient of determination (R²) score of 0.75 and a root-mean-square error (RMSE) of 1.35 m on the test dataset. These findings indicate that the combination of SHAP-driven multi-category feature fusion and deep learning can effectively enhance both the inversion accuracy and physical interpretability, providing a reliable solution for PolSAR-based forest structural parameter retrieval at the Huangfengqiao study site, with potential applicability to complex terrain conditions. Full article

Accurate identification and spatial localization of hidden hazards in embankments are essential for the reinforcement and safety management of defective structures. However, conventional drilling and single geophysical methods are often insufficient for fine-scale detection due to the strong heterogeneity of embankment materials, complex internal structures, and diverse forms of leakage-related defects. To address these challenges, this study establishes conceptual models for two representative embankment types, namely homogeneous embankments and core-wall embankments, based on reflection-wave imaging theory. The propagation characteristics and imaging responses of reflection waves in embankment media are systematically investigated. A forward-modeling approach based on the shortest-path ray tracing method is developed, and reflection-wave imaging is achieved through travel-time tomography inversion. The diagnostic results show that the proposed reflection-wave imaging method can effectively delineate the spatial distribution and geometric morphology of internal defects, demonstrating strong capability in identifying leakage channels and loose zones. The research provides a theoretical basis and technical support for nondestructive detection and comprehensive diagnosis of embankment hazards. Full article

Background: Young adulthood represents a critical period for the emergence of early metabolic disturbances, potentially influenced by dietary shifts toward convenience and ultra-processed foods. However, evidence linking dietary patterns and cooking practices with objective metabolic biomarkers in Romanian university students remains limited. Methods: This cross-sectional study included 693 students aged 18–24 years at the Victor Babeș University of Medicine and Pharmacy, Romania (June–July 2025). Dietary habits, food preferences, and cooking practices were assessed using a structured online questionnaire, while anthropometric and biochemical data were obtained from university health records. The primary outcome was glycated hemoglobin (HbA1c), a marker of average blood glucose levels over the previous 2–3 months. Prediabetes was defined as HbA1c 5.7–6.4%. Dietary patterns were identified using k-means clustering based on fast-food consumption frequency, main meal of the day, fruit and vegetable intake frequency, and predominant cooking method. Multivariable regression models assessed associations between dietary variables and glycemic or lipid outcomes. Results: Prediabetes prevalence was 21.1% (diabetes: 1.4%). Three dietary patterns were identified: health-conscious (prediabetes 15.4%), mixed (20.0%), and fast-food oriented (27.3%; χ² p = 0.003). Fast-food consumption frequency was independently associated with higher prediabetes risk (OR = 1.78 per category; 95% CI 1.38–2.30; p < 0.001) and higher HbA1c levels (β = 0.147; p < 0.001), while fruit and vegetable intake showed an inverse association with HbA1c (β = −0.109; p < 0.001). A dose–response relationship was observed between fast-food frequency and both HbA1c and prediabetes prevalence (p-trend < 0.001). An interaction between high-temperature cooking methods and frequent fast-food consumption was observed for HbA1c (p = 0.023). BMI and sex were the strongest predictors of lipid outcomes, although fast-food intake was associated with higher triglyceride levels (p = 0.034). Conclusions: Among Romanian university students, dietary patterns characterized by frequent fast-food consumption were associated with higher HbA1c levels and greater prediabetes prevalence. A high-temperature cooking method was associated with higher glycemic levels when combined with frequent fast-food intake. These findings suggest that early dietary behaviors during university years may be relevant for metabolic risk profiles in young adults. Full article

Background: Long COVID is characterised by persistent symptoms after SARS-CoV-2 infection, and its impact on cardiovascular health is a growing concern. This study aimed to evaluate the association between the presence and severity of metabolic syndrome and vascular structural and functional in patients with long COVID. Methods: We conducted a cross-sectional study of 304 adults diagnosed with long COVID. Vascular health was assessed using carotid intima–media thickness to evaluate arterial structure, and pulse wave velocity to assess arterial stiffness. Metabolic syndrome was defined according to international criteria. Multiple regression models were performed to analyse the association between the number of metabolic syndrome components and vascular parameters, adjusting for age, sex, lifestyle and pharmacological treatments. Results: All vascular measures show a positive association with artery pressure. All measures except cardio–ankle vascular index were positively associated with the number of metabolic syndrome components. Carotid intima–media thickness, carotid–femoral pulse wave velocity and vascular ageing index were positively associated with waist circumference. Brachial–ankle pulse wave was positively associated with all metabolic syndrome components and showed an inverse association with HDL-cholesterol. Cardio–ankle vascular index was inversely associated with waist circumference. Conclusions: In conclusion, among adults with long COVID, metabolic syndrome and the accumulation of its components are associated with poorer vascular structure, function, and vascular ageing. Full article

Wafer map defect pattern recognition is a key task for ensuring yield in integrated circuit manufacturing. However, in real production lines it commonly suffers from scarce labeled data, long-tailed class distributions, and limited feature representations, which cause existing deep learning models to degrade in performance, particularly for minority defect classes and complex defect morphologies. To address these challenges, we propose a semi-supervised classification method for wafer maps based on a multimodal Wasserstein autoencoder (MM-WAE). The framework constructs three parallel feature branches in the spatial, frequency, and texture domains, using a multi-head attention mechanism and gating mechanism for adaptive multimodal fusion. This allows defect patterns to be comprehensively characterized by macroscopic geometric distributions, spectral periodic structures, and microscopic texture details. The Wasserstein autoencoder is introduced, with the latent space distribution regularized by a maximum mean discrepancy (MMD) loss using an inverse multiquadratic kernel. Additionally, an inverse class-frequency weighted cross-entropy loss and a modality consistency loss between the encoder and classifier jointly optimize the reconstruction and classification paths while leveraging large amounts of unlabeled wafer maps for semi-supervised learning. Experimental results show that MM-WAE mitigates performance limitations caused by insufficient labels and class imbalance, significantly improving the accuracy and robustness of wafer defect classification, with promising potential for industrial application and further development. Full article

Satellite remote sensing increasingly supports water quality monitoring, yet the temporal transferability of machine learning (ML) models remains insufficiently tested, particularly in coastal shallow lakes subject to hydrological variability. This study evaluates the predictive robustness of satellite-based ML models for electrical conductivity (EC), turbidity (TUR), water temperature (WT), and dissolved oxygen (DO) in Vrana Lake, Croatia. A total of 409 in situ measurements collected during 2023–2024 and 2025 were paired with Sentinel-2 and Landsat 8–9 imagery. Pearson, Spearman, and Kendall correlation analyses were applied for parameter-specific band selection using original, inverse, quadratic, and logarithmic feature transformations. Seventeen regression algorithms were evaluated under six training–testing split strategies, including strict temporal projection. WT exhibited high robustness (R² ≈ 0.90 under temporal projection) due to its strong dependence on thermal bands, while DO achieved moderate temporal stability (R² = 0.51) using log-transformed predictors. EC and TUR demonstrated substantial performance degradation under temporal separation (R² = 0.14 and −4.62, respectively), reflecting sensitivity to distribution shifts. For parameters showing sufficient stability, interpretable band-based retrieval equations were derived using the most strongly correlated spectral predictors. These findings highlight the importance of temporally structured validation and demonstrate that model complexity does not guarantee operational robustness in shallow, dynamically evolving lake systems. Full article

Real-time mechanical analysis within digital twin (DT) systems requires high-fidelity models that synchronize with the “as-built” state of physical structures. This paper proposes a technical framework for constructing a “mechanical-core” DT by integrating computer vision (CV) sensing with automated finite element model (FEM) updating. Utilizing the Midas API, we developed a platform that automates data acquisition, modeling, and parameter inversion. A momentum-based optimization algorithm is implemented to invert the instantaneous elastic modulus of bridge segments during cantilever construction. The system was validated through a case study of a continuous box-girder bridge. Quantitative results indicate that the initial theoretical model, based on design-phase assumptions, exhibited a mean relative error of approximately 21.9% in vertical displacement. Following the real-time parameter inversion, this error was significantly reduced to less than 0.2% across all monitored construction stages. The rapid convergence (typically within three iterations) and the substantial increase in predictive accuracy demonstrate that the proposed framework effectively bridges the gap between raw sensing data and structural analysis, providing a reliable basis for proactive engineering decision-making. Full article

Suspended sediment concentration (SSC) strongly influences estuarine erosion–deposition processes, navigation safety, and coastal engineering stability. However, conventional remote sensing techniques are limited to surface SSC and cannot characterize vertical sediment structures. In this study, Landsat 8 OLI imagery was combined with in situ SSC profiles from six stations in the Guan River Estuary–Haizhou Bay system to retrieve full-depth sediment distributions. A band-combination inversion model using (B3 + B2)/B1 achieved the highest accuracy (R² = 0.679), and an improved vertical distribution model was developed by incorporating turbulent shear (G) into the Rouse framework. Results indicate that surface SSC ranged from 0.15 to 0.86 kg/m³, while middle- and bottom-layer SSC reached up to 1.20 kg/m³ and 1.77 kg/m³, respectively, exhibiting a consistent east–high and west–low spatial pattern. Settling velocity (SSV) varied from 3 × 10⁻⁶ to 1.49 × 10⁻² m/s and showed a positive correlation with SSC at low concentrations and a negative correlation at high concentrations due to flocculation effects. This integrated framework provides a rapid, low-cost method for full-water-column sediment assessment in estuaries and coastal zones, supporting engineering design, navigation maintenance, and sediment management. A better understanding of sediment transport processes in Haizhou Bay is important for maintaining shoreline stability and ecological balance in this semi-enclosed coastal system. The findings of this study provide a scientific basis for sediment management and environmental regulation, which can contribute to the long-term sustainable development of coastal environments in the Yellow Sea region. Full article

In autonomous driving, deep convolutional neural networks face a core conflict between computational efficiency and safety-critical robustness on resource-constrained onboard computing units. Dominant structural pruning, based on weight magnitude or geometric statistics, fails in long-tailed traffic scenarios by equating parameter magnitude with feature importance and pruning critical filters in the tail classes. To address this, we propose a structural pruning framework that evaluates the semantic utility of features using weighted copula entropy rather than relying solely on their magnitude. Our novel approach integrates Elastic Net regularization for inducing sparsity and weighted copula entropy for unbiased information-theoretic feature selection. By incorporating inverse class frequency weighting into empirical Copula estimation, we decouple feature relevance from sample abundance, ensuring the preservation of rare-class discriminators based on their information content rather than occurrence frequency. Furthermore, this metric is embedded into an enhanced max-relevance and min-redundancy algorithm to eliminate semantic redundancy while maintaining representational diversity. Extensive experiments on the BDD100K dataset with YOLOv5l and YOLOv8l architectures demonstrate that, at a 50% pruning rate, the proposed method reduces FLOPs and parameters by nearly 50%, with only 0.09% mAP@0.5 loss for YOLOv5l and 0.14% mAP@0.5 loss for YOLOv8l, while significantly improving the mAP of the extreme tail class Train from 0% to 3.84% and 2.76% to 5.12%, respectively. It achieves a more favorable trade-off between detection accuracy and computational efficiency than mainstream pruning approaches. This work provides a lightweight scheme for autonomous driving perception models and a new information-theoretic perspective for structured network pruning. Full article

Biological cardiovascular vulnerability is defined as an imaging-derived construct integrating myocardial functional impairment, coronary microvascular dysfunction, and modeled hemodynamic burden, including global longitudinal strain, coronary flow reserve, and derived vascular indices. To evaluate whether advanced echocardiographic and coronary Doppler imaging parameters identify biological cardiovascular vulnerability associated with the severity and complications of infective endocarditis beyond conventional structural findings. In this retrospective single-center cohort study, we analyzed consecutive patients with definite infective endocarditis who underwent advanced echocardiographic and coronary Doppler imaging. Comprehensive transthoracic and transesophageal echocardiography assessed vegetation characteristics, left ventricular function, global longitudinal strain (GLS), diastolic indices, right ventricular function, and pulmonary artery systolic pressure. Coronary microvascular function was evaluated noninvasively using transthoracic Doppler-derived coronary flow reserve (CFR) of the left anterior descending artery. Associations with disease severity and perivalvular complications were evaluated using multivariable regression analysis. Reduced coronary flow reserve was independently associated with the composite severe infective endocarditis phenotype, as defined by perivalvular complications, severe valvular dysfunction, or endocarditis team-guided urgent surgical indication. Coronary flow reserve correlated inversely with vegetation size (r = −0.39; p = 0.002) and regurgitation severity (r = −0.36; p = 0.004). Notably, the inverse association between coronary flow reserve and vegetation size showed substantial interindividual variability, particularly among patients with similar vegetation dimensions, suggesting heterogeneity in microvascular vulnerability beyond structural lesion burden. Despite relatively preserved mean arterial pressure across age groups, advanced imaging revealed progressive increases in systemic vascular resistance, declining wall shear stress, impaired microvascular flow, and reduced myocardial reserve. Imaging-derived cardiovascular vulnerability profiles frequently diverged from chronological age, highlighting heterogeneity in cardiovascular reserve despite apparently stable conventional hemodynamic parameters. Advanced echocardiographic and coronary Doppler imaging characterize a spectrum of biological cardiovascular vulnerability that is associated with clinically adjudicated severity in infective endocarditis, rather than serving as independent prognostic predictors. Full article

Parameter estimation for biochemical reaction networks is computationally demanding, especially for systems with oscillatory nonlinear dynamics, where standard iterative optimization strategies, including genetic algorithms, often struggle with prohibitive computational costs. We introduce an efficient parameter estimation framework that combines a real-coded genetic algorithm with a novel adaptive simulation termination strategy. This strategy defines a time-dependent termination boundary based on population quantiles, which is permissive during early transients and becomes progressively stricter as simulations advance, explicitly accounting for the temporal structure of oscillatory behavior. Crucially, this mechanism facilitates the efficient identification and early simulation termination of poor parameter candidates, thus avoiding the computational expense of full-horizon simulations. The framework further integrates global exploration with the modified Powell method for rapid local refinement. Numerical experiments on two benchmark oscillatory models—the Lotka–Volterra and Goodwin oscillators—demonstrate that the framework reduces computational cost by approximately 30–50% compared to a baseline GA without this strategy. For the parameter-sensitive Goodwin model, the framework efficiently identifies candidates evolving toward damped oscillations caused by subtle parameter variations. Sensitivity analysis also confirms robustness across diverse hyperparameter settings, indicating that adaptive simulation termination provides a practical acceleration mechanism for inverse problems in systems biology where iterative objective function evaluation dominates runtime. Full article

This study proposes a theoretical–empirical framework for analyzing political regimes based on a structural analogy between electoral behavior and spin-glass systems in statistical physics. Society is modeled as a system of interacting agents (voters) influenced by both interpersonal interactions and external factors such as media and institutions, formalized through a social Hamiltonian. By introducing a partition function and free energy, political regimes are interpreted as distinct macroscopic phases governed by four effective macro-parameters: external field, conformism, interaction heterogeneity, and inverse social temperature. Democratic societies correspond to a multistable regime characterized by sensitivity to initial conditions and replica symmetry breaking (RSB), reflecting the coexistence of competing social configurations. Authoritarian regimes, in contrast, arise when a strong unidirectional external field, high conformism, and low effective social temperature stabilize a single dominant macroscopic state, producing a regime analogous to replica symmetry (RS). A central result of the model is the distinction between the predictability of macroscopic outcomes and structural social multistability, as well as between natural and externally imposed homogenization of collective behavior. To illustrate the empirical relevance of the framework, the model is applied to the transition from the Weimar Republic to the National Socialist regime (1919–1933), using aggregated electoral data to construct proxy indicators for the effective parameters governing social interactions. The proposed approach enables structural identification of early signals of authoritarian transition through changes in the parameters of social dynamics. Full article

Deep learning approaches for egocentric video understanding often lack a principled theoretical treatment of stability, particularly when dealing with the sparse, noisy, and temporally ambiguous observations characteristic of first-person imaging. In this work, we frame egocentric video question answering not merely as a classification task, but as an ill-posed inverse problem aimed at reconstructing latent semantic intent from stochastically perturbed visual signals. To address the instability inherent in standard dual-encoder architectures, we present a framework with a mathematical interpretation that incorporates gated cross-modal interaction within the transformer backbone. Formally, the video-side update analyzed in this work is defined as a learnable convex combination of unimodal feature representations and cross-modal attention residuals; the full implementation applies analogous gated cross-modal updates bidirectionally. From a regularization perspective, the gating mechanism can be interpreted as an adaptive parameter that balances data fidelity against language-conditioned structural constraints during feature reconstruction. We provide the Bounded Update Property (Lemma 1) and an analytical layer-wise sensitivity bound and empirically demonstrate that the proposed framework achieves measurable improvements in both accuracy and stability on the EgoTaskQA and MSR-VTT benchmarks. On EgoTaskQA, our model improves accuracy from 27.0% to 31.7% (+4.7 pp) and reduces the accuracy drop under 50% frame drop from 3.93 pp to 0.94 pp. On MSR-VTT, our model improves accuracy by 13.0 pp over the dual-encoder baseline. Under severe perturbation (50% frame drop) on MSR-VTT, our model retains 97.7% of its clean performance, whereas the baseline exhibits near-zero drop accompanied by majority-class behavior. These results provide empirical evidence that the proposed interaction induces stable behavior under perturbations in an ill-posed multimodal inference setting, mitigating sensitivity to sampling variability while preserving query-relevant temporal structure. Furthermore, an entropy-based analysis indicates that the gating mechanism prevents excessive diffusion of attention, promoting coherent temporal reasoning. Overall, this work offers a mathematically informed perspective on designing interaction mechanisms for stable multimodal systems, with a focus on robust reasoning under temporal ambiguity. Full article