Concrete acceptance continues to rely almost exclusively on compressive strength, yet deterioration-driven failures under aggressive exposure, such as reinforcement corrosion and freeze-thaw attack, remain a leading cause of premature decommissioning, even for strength-compliant members. This mismatch points to a fundamental gap: in-service performance is governed not by any single metric but by the joint risk of non-compliance across coupled strength, durability and denseness indicators, whose contradictions the current pass/fail paradigm systematically overlooks. Addressing this gap, the present study develops a data-driven multi-indicator quality assessment workflow that chains point prediction, distribution-free uncertainty quantification and taildependent joint non-compliance modelling into an integrated assessment pipeline. Eight machine-learning models are benchmarked for simultaneous prediction of compressive strength, electrical resistivity and ultrasonic pulse velocity on 4420 specimens with synchronous measurements; the reported accuracy (R2 = 0.984-0.989 for compressive strength) strictly reflects in-domain interpolation, as leave-one-group-out crossvalidation by mix design yields negative R2 for key indicators. Split conformal prediction then furnishes finitesample prediction intervals whose coverage guarantees hold without distributional assumptions, while grouped analyses expose pronounced coverage heterogeneity across curing ages and aggregate types. A trivariate t-copula fitted to the prediction residuals reveals that the dependence among indicators is concentrated in the extreme tails: when compressive-strength residuals fall in the worst 5% region, the conditional probability of concurrent resistivity deterioration is roughly sevenfold higher than the independence baseline (an order-of-magnitude estimate given the small tail sample), a finding that reframes "strength-compliant yet durability-deficient" cases as a robustness deficit of the acceptance system itself. Joint non-compliance probabilities estimated via Monte Carlo simulation provide a graded quality-screening map that can inform acceptance decisions and lifecycle quality management workflows. The framework is modular: base model, conformal strategy and copula family are each independently replaceable. It is also, in principle, readily transferable to broader multi-indicator quality assessment problems in construction materials and infrastructure systems.
Next Point-of-Interest (Next-POI) recommendation predicts a user's next location based on historical check-in sequences, requiring effective modeling of temporal dynamics, spatial dependencies, and a balance between long-and short-term preferences. Existing methods often conflate spatio-temporal signals and rely heavily on Transformer-based encoders, which can cause over-smoothing and overlook fine-grained spatio-temporal structures. Moreover, traditional methods mostly adopt a single-view contrastive learning mechanism, which limits their ability to comprehensively capture diverse behavior patterns. Separately, many models depend on single-model ranking outputs and lack flexible fusion mechanisms to combine strengths from multiple ranking signals. To overcome these limitations, we propose FMMCRec, a frequency-aware framework for Next-POI recommendation which integrates a Dual Positional Encoding Decoupled Transformer with Polynomial Spectral Decomposition in Frequency domains to disentangle temporal and spatial signals while reducing representation homogenization associated with over-smoothing. A lightweight spectral candidate generator improves POI selection quality and robustness. Besides, a Multi-stage re-ranking module flexibly fuses ranking signals via Borda count and hit-rate-aware scoring. For representation learning, a Multi-view Contrastive strategy captures diverse behavior patterns with category-aware factors. To support semantic generalization to user-unvisited, category-related POIs, we introduce a category-aware positive sampling strategy for contrastive augmentation. Extensive experiments on three real-world datasets show that FMMCRec improves recommendation accuracy over state-ofthe-art baselines, with additional diagnostic evidence for representation diversity under the evaluated settings. The code is available at https://github.com/MyCode12-maker/FMMC.
In this paper, we propose an asymmetric fiber-optic Mach-Zehnder interferometer based on multimode fiber (MMF) and no-core fiber (NCF) for vector curvature sensing. The proposed sensor is fabricated by core-offset splicing using arc discharge and then CO2 laser polishing of the core-offset region. The beam propagation method is employed to simulate the influence of different structural parameters on the transmission spectrum, thereby determining the optimal parameters. Based on these simulation results, experimental validation was conducted, which demonstrates that the proposed structure can distinguish between curvature applied along the + X, −X and Y directions with the maximum curvature sensitivities of −29.0 nm/m−1, 10.9 nm/m−1 and 0.8 nm/m−1, respectively. The different wavelength responses of the sensor under different bending directions arise from the asymmetry caused by core-offset and side polishing. With its high sensitivity, easy fabrication, and direction discrimination, the proposed sensor is promising for bridge health monitoring and aerospace material inspection.
Alkali metal poisoning of NH3 selective catalytic reduction (NH3-SCR) catalysts remains a critical bottleneck in NOx control for stationary pollution sources. In this study, a hollow Mn0.1TiOx nanotube (MT-T) catalyst was constructed through an in-situ co-growth strategy, enabling precise tuning of its microstructure and surface properties. Compared with conventional nanoparticles (MT-N), MT-T exhibited a larger specific surface area (280 m2/g), a higher density of acid sites, and superior structural stability, achieving nearly 100% NO conversion and high N2 selectivity over a wide temperature range of 150-390 degrees C. Even after 2 wt% K poisoning, MT-T maintained excellent catalytic activity, demonstrating remarkable resistance to alkali metal poisoning. H2TPR, NH3-TPD, XPS, and other characterization results indicated that K poisoning typically induces structural changes in the catalyst, loss of acid sites, and perturbation of its electronic environment, whereas in MT-T, the cooperative effect between the hollow nanotube structure and the strong K-Ti interaction markedly mitigates these adverse effects, thereby better preserving its active sites, which was helpful for achieving excellent SCR performance.
Spin polarization of metal active centers provides a powerful means to mitigate the spin flipping of key intermediates during electrocatalysis. However, effectively triggering spin polarization and establishing its relationship with performance in oxygen reduction/evolution reactions (ORR/OER) remain challenging. This study proposes a defect engineering means that creates selenium vacancies (SeV) in cobalt diselenide (CoSe2) integrated onto an iron single-atom platform (FeSA@CoSe2-SeV) to enhance Co's spin polarization. FeSA@CoSe2-SeV achieves a high ORR half-wave potential (0.921 V) and a low OER overpotential (370 mV@10 mA cm-2), significantly outperforming FeSA@CoSe2. When applied to zinc-air battery (ZAB), it achieves a high power-density (186 mW cm-2), with a cycling life of up to 582 h. FeSA@CoSe2-SeV-based flexible ZAB maintains stable charge/discharge performance even under 0 degrees-180 degrees bending conditions. Introduction of SeV reduces the degeneracy of the Co 3d orbitals, effectively triggering spin polarization. This electronic structure reconstruction causes the pi* orbital of Co-*O/*OH to lose an electron, enhancing the hybridization between Co 3d and *O/*OH 2p orbitals and thereby mitigating the intermediates' spin flipping. Ferromagnetic FeSA stabilizes the SeV and Co active sites, ensuring the structural/catalytic stability. This work confirms the effectiveness of SeV-induced spin polarization regulation, providing a novel spintronics-based approach for designing bifunctional electrocatalysts.
