Search Results (1,205)

Finite element (FE) analysis that incorporates structural damages extracted using computer vision (CV) methods based on three-dimensional point-cloud data (PCD) is effective for evaluating the residual capacity of structures. In this study, observable surface damages are regarded as surface anomaly regions on the structure. Based on this assumption, a modular mesh-model updating framework is proposed to incorporate damage information into shell-element FE models through the anomaly detection from point-cloud data. The framework comprises three modules: a perception module for the anomaly detection and anomaly-region localization, a description module for the anomaly-region quantification, and a remeshing module for mapping the anomaly region onto the FE model. Its performance was evaluated using a steel angle member specimen with artificially introduced anomaly regions representing damage. The updating results were evaluated in terms of both geometric accuracy and FE analysis (FEA) applicability by comparison with the reference FE model manually constructed from the designed anomaly-region geometries. The Intersection over Union (IoU) between corresponding anomaly regions in the updated and reference models ranges from 0.44 to 0.87, including the individual plate results for cross-surface cases. Through static elastoplastic analysis, the predicted reductions in ultimate load-bearing capacity differed by approximately 1.2 percentage points between the two updated models. Moreover, local stress redistribution showed qualitative agreement, although quantitative discrepancies remained in the magnitudes and locations of stress concentrations. Together with several supporting components, the proposed framework provides an effective workflow for integrating PCD-based CV techniques into FE model updating and has potential for FEA-based damage assessment of structures. Full article

The efficient expression of recombinant proteins in aquaculture fish poses a significant challenge, partly due to the lack of dedicated bioinformatics tools for species-specific codon usage analysis and optimization. RiboScan (v1.0.0), a portable, single-file command-line tool executed in Python, was designed in this study, for mRNA codon diagnostics and optimization. RiboScan supports ten species, including five major aquaculture fish. High-risk windows that are computationally predicted to impair translational efficiency are identified based on two complementary metrics, namely, the Relative Adaptiveness Index (RAI) and estimated NN-based stability score, which are calculated using a sliding-window approach. An optimization module is used to replace the suboptimal codons in high-risk regions with host-specific synonymous codons, which are selected based on their high RAI values, leaving the encoded protein sequence unaltered. The efficacy of RiboScan was evaluated using coding sequences (CDSs) from three representative aquaculture fish genes, namely, Salmo salar growth hormone (gh; GenBank accession: M21573), Oreochromis niloticus interleukin-1 beta (il1b; GenBank accession: OR432591), and Ctenopharyngodon idella beta-actin (actb; GenBank accession: M25013). The tool identified 25–30 high-risk windows per CDS pre-optimization, whereas no high-risk windows were detected in any of the CDSs post-optimization. The mean RAI for Sal. salar gh, Or. niloticus il1b, and Ct. idella actb increased from 0.189 to 0.339, 0.126 to 0.270, and 0.140 to 0.231, respectively. RiboScan’s main analytical workflow requires no third-party dependencies, as it uses only the Python standard library. In addition, the tool yields publication-ready figures and structured comma-separated values (CSV) output files. RiboScan can be used to perform species-specific codon optimization for major aquaculture fish, and provides a practical computational framework for designing recombinant proteins, vaccines, and other engineered genetic constructs, thus contributing to aquaculture biotechnology research. Full article

Pepper suffers from bacterial leaf spot and yellow leaf curl, which substantially reduce crop yield and fruit quality. Traditional manual diagnosis suffers from delayed response, subjective bias and heavy labor consumption, while existing intelligent detection pipelines lack standardized preprocessing workflows, quantitative feature screening and systematic model comparison. To fill these research gaps, we built a pepper leaf dataset with 1260 samples (healthy, bacterial spot, yellow leaf curl). Three segmentation algorithms (Lab b-channel, RGB super-green, Otsu-ACWE) were quantitatively assessed to select the optimal preprocessing scheme. We extracted 32 fused visual features (27 RGB/HSV/Lab color moments + five gray-level co-occurrence matrix (GLCM) texture metrics) and adopted a random-forest classifier to eliminate seven low-contribution redundant features, retaining 25 discriminative variables. Three representative models, namely convolutional neural network (CNN), logistic regression (LR), and genetic-algorithm-optimized back-propagation neural network (GA-BP), were constructed for parallel comparison via 20 independent repeated trials, with accuracy, precision, recall, F1-score and area under the receiver operating characteristic curve (AUC) as evaluation indicators. The results verified that Lab b-channel segmentation achieved superior background separation and intact lesion edge retention. CNN yielded the best performance, with an average test accuracy of 97.67% and an average AUC of 0.999, accompanied by minimal metric standard deviations and outstanding stability. LR exhibits low computational cost and fast training, which is promising for applications with limited computing resources. In contrast, GA-BP shows weak nonlinear fitting ability and severe prediction fluctuations, making it unsuitable for high-precision diagnosis. This study proposes a standardized experimental framework to offer theoretical guidance and algorithmic references for intelligent vegetable leaf disease identification. All experiments were conducted on a dataset collected under standardized indoor single-illumination conditions; therefore, the conclusions of this study are only applicable to such controlled scenarios. Full article

The Huaihe River Basin (HRB) lies in the climatic transition zone between northern and southern China. Its precipitation shows obvious spatiotemporal heterogeneity, and frequent meteorological droughts seriously threaten regional food and water security. Clarifying the spatiotemporal variations, non-linear abrupt changes and multi-scale driving mechanisms of meteorological droughts in the basin is of great significance for regional drought risk prevention and control, as well as the optimized allocation of water resources. In this study, the one-month time scale Standardized Precipitation Evapotranspiration Index (SPEI-1) was adopted as the primary indicator for quantifying meteorological drought. An analytical workflow integrating Bayesian Estimator of Abrupt Change, Seasonality, and Trend (BEAST), MMK–Hurst coupling trend and persistence discrimination, Three-Threshold Run Theory and Partial Wavelet Coherence (PWC) is constructed. The framework systematically investigates drought spatiotemporal evolution, abrupt change features, persistent trend patterns and climatic driving effects over 1982–2024. The results indicate the following: (1) In 1982–2024, the drought in the whole basin showed a slight aggravating trend, with an average drought trend rate of −0.000387. Spatially, this trend varied, with faster progression in the west and slower in the east, and more severe conditions in the west and milder conditions in the east. (2) 1988 was the driest year within the study period, with three drought peaks occurring in April, June and November. Annual mean SPEI-1 values indicated the severest drought in the Yishu-Si River system (YSR, −0.62), followed by the Shandong Peninsula and Coastal River systems (SPCR, −0.56), Huai River Mainstream River system (HRMR, −0.55), and the Lixia River system (LR, −0.52). (3) The probability that the mutation point for the seasonal component of the SPEI occurred in March 2001 was 74%, whilst the probability that the potential mutation signal for the trend component occurred in September 1998 was 44.1%. (4) Within the HRB, droughts covering over 95% of the basin intensified in spring and autumn. A mean Hurst index of 0.70 implies overall persistent drought evolution. MMK–Hurst coupled analysis revealed that the basin was predominantly dominated by mild, non-significant, persistent drought. (5) The typical cross-seasonal drought event of 1998–1999 exhibited multi-stage fluctuations, with the central and western hilly regions constituting the core cluster of extreme droughts. During this event, areas experiencing moderate drought accounted for 41.79%, whilst those experiencing extreme drought accounted for only 1.47%, and drought intensity diminished progressively from west to east. (6) Air-specific humidity (AH) constitutes an Average Wavelet Coherence (AWC) of 0.94 and a Percentage of Significant Power (POSP) of 12.20%. AH achieves the highest total POSP with only a marginal advantage relative to SM. The multi-method coupled analysis framework established in this study provides theoretical support for regionalized drought early warning, water resource regulation, and disaster prevention and mitigation in the HRB. Full article

Smart-building structural health monitoring (SHM) requires a unified digital representation capable of integrating heterogeneous sensing devices, continuous structural states, and burst-oriented post-event assessment without embedding device-specific logic throughout the software stack. This study proposes a semantic digital twin architecture in which SensorType, DeviceProfile, and site metadata form a semantic single source of truth and generate W3C Web of Things Thing Descriptions at runtime. The resulting WoT-driven contract governs field mapping, schema-on-write persistence, generic API access, state visualization, and engineering-threshold evaluation. To accommodate heterogeneous temporal behavior, event-driven seismic assessment and state-driven construction tilt monitoring are orchestrated as distinct workflows that share persistence, notification, and observability services while retaining separate timing contracts. Controlled extension experiments required no manual data-layer, backend, ingestion, or frontend modification, with a runtime source-hash difference of zero. Under a ten-building seismic-event burst, continuous write-lag p95 changed by $-20$ ms from a 969 ms baseline while all event jobs completed without restart or out-of-memory conditions. The ingestion path further sustained 71,040 points/s at 300 sensors with no dropped points. These results demonstrate that WoT-driven semantic interoperability and event–state workflow orchestration can provide an extensible integration foundation for smart-building SHM within a clearly defined configuration boundary. Full article

The Industry 4.0 transition of construction requires evidence linking digital maturity to sustainability outcomes in operations management. Yet Building Information Modelling (BIM) maturity is still assessed through a priori stage models rather than the adoption patterns companies actually exhibit, and its co-occurrence with perceived sustainability performance is uncharacterised beyond single-country samples. This paper derives a typology of lifecycle BIM-use maturity in Central European construction companies and characterises how the types differ in perceived sustainability performance. Four objectives are pursued: the indicators’ dimensional structure and reliability; derivation of the typology by unsupervised learning and the support for its boundaries; each type’s sustainability profile, national composition, and incremental information beyond country and firm scale; and delimitation of the BIM–sustainability association. A structured questionnaire administered to 199 companies in Croatia, Slovakia and Slovenia over five years yielded six lifecycle BIM-use items and three perceived-sustainability items (recycling, waste, CO₂), which were analysed by reliability assessment, principal component analysis, and k-means and Ward clustering with country-stratified checks. Two dimensions—adoption intensity and an end-of-life-versus-design orientation—accounted for 78.1% of the variance. The 199 independently surveyed companies exhibit only seven distinct response profiles, which bounds the typology’s resolution; no internal validity index showed an interior optimum, so the four-type solution is chosen rather than validated. The types differ systematically in perceived sustainability; the gradient holds within Croatia (with one inversion) and Slovakia but reverses in Slovenia. The typology is exploratory and perception-based—a precursor to objective lifecycle assessment rather than a substitute: the BIM-use and sustainability items are empirically proximate and failed a common-method-variance check, so their relation is descriptive co-occurrence. One result is independent of that caveat: substantial end-of-life BIM use appears in only one of the seven profiles, so the model data that circular workflows require are largely absent. Full article

Modern building infrastructures are becoming increasingly complex, creating a need for intuitive indoor navigation systems that can assist users in unfamiliar environments. Augmented Reality (AR) has emerged as a promising solution by providing spatially contextual guidance directly within the user’s field of view. However, many AR indoor navigation systems rely on manually constructed 3D environments, a development process that is time-consuming and prone to spatial inconsistencies with the real-world environment. This study presents a comparative evaluation of two environment creation workflows for AR indoor navigation development: a traditional manual 3D modeling approach and an automated cloud-based spatial mapping workflow using the Immersal SDK. A counterbalanced within-subject experiment was conducted with 48 participants, each of whom completed equivalent indoor navigation development tasks using both workflows in a real-world campus building environment. The development process was divided into three stages: environment acquisition, environment generation, and system integration. Development efficiency was evaluated using stage-based development time measurements, while perceived workload was assessed using the NASA Task Load Index (NASA-TLX). Statistical analysis was performed using repeated-measures analysis to compare workflow performance across development stages. Results show that the automated workflow significantly reduced overall development time by approximately 38% compared to the manual modeling approach, with the most substantial time reductions occurring during the environment acquisition and environment generation stages. NASA-TLX results indicate an approximately 31% reduction in overall perceived workload. Descriptively, the automated workflow had lower mental-demand and effort scores but a higher physical-demand score. A separate researcher-conducted spatial validation of one implementation per workflow showed a higher mean three-dimensional positional error for the automated implementation (22.47 cm) than for the manual implementation (19.14 cm), with a mean paired difference of 3.33 cm across 13 anchor locations. These findings indicate that automated spatial mapping can substantially improve development efficiency and reduce overall perceived workload, while introducing trade-offs in physical demand and spatial alignment accuracy relative to manual environment reconstruction. Full article

Gel electrophoresis is widely used for protein and nucleic acid analysis. Nucleic acid recovery from agarose gels is convenient and efficient, but protein recovery is not. This study aimed to develop and validate a novel method for recovering proteins from gels. Methods: After protein separation on gels, excised gel slices containing target proteins are processed using the same silica-column workflow as that used for nucleic acid purification, but with buffers specifically adapted for protein recovery. This approach only requires brief centrifugation steps and avoids laborious extraction and specialized equipment. Results: Proteome-wide recovery showed a slight preference of the method for low-molecular-mass and acidic proteins, while individual protein recovery efficiencies ranged from 69.9% to 86.7% and showed no correlation with intrinsic protein properties. The method effectively supported antigen purification. A banana (Musa acuminata AAA group ‘Brazilian’) proteome library constructed from recovered fractions contained over 18,000 protein groups and 27,000 individual proteins, which significantly enhanced protein identification in library-based data-independent acquisition (DIA) compared to directDIA analysis of single samples. Conclusions: This method bridges gel-based protein separation- and column-based purification, making protein purification as simple, fast, and scalable as nucleic acid purification. It offers a practical alternative for proteomic sample preparation, antigen purification, and DIA workflows. Full article

Embodied-CO₂ savings in clinker-reduced concrete must be evaluated alongside cover-zone durability and service life. This review examines carbonation and shrinkage/restrained cracking as coupled, rather than independent, controls on atmospheric service life in reinforced concrete buildings. The synthesis links binder route to cover-zone transport and alkalinity buffering and explains how curing, preconditioning, relative humidity (RH), CO₂ concentration, exposure geometry, and crack state affect reported carbonation rankings. It develops a two-domain conceptual framework for interpreting the effective carbonation front in cracked cover, with a descriptive crack-amplification index for reporting localized penetration. Practical mitigation is organized around mix-design levers, curing and construction actions, and a three-tier performance-based assessment workflow with a coupled test matrix. The review closes with a reporting checklist and five research gaps. The central conclusion is conditional: clinker-reduced building concrete can meet specified carbonation-performance requirements when curing, RH history, paste volume, and crack connectivity are assessed and managed together. Full article

Frequency-controlled parallel inverters combine nonlinear switching dynamics with load-dependent periodic operation, which makes controller-oriented uncertainty modeling difficult. This paper presents a reproducible surrogate-to-robust-control workflow. A local slope calibration maps normalized switching-frequency variation to an equivalent continuous input, and four physical resistance–inductance corner linearizations are fitted by a three-state model with two real coefficient-space uncertainties. A third-order fixed-structure controller is obtained by finite-budget minimax mixed-sensitivity optimization and compared with a nominal filtered proportional–integral controller. Independent verification on a 31 × 31 uncertainty grid and a continuous worst-case search gives a peak-weighted gain of 0.028920 versus 0.089893 for the comparison controller, a 67.83% reduction in the adopted criterion; all 500 Monte Carlo closed loops are stable. Interior Jacobian errors remain below 0.010%, and a residual-augmented coefficient box preserves the robust-performance conclusion. Cycle-to-cycle nonlinear simulations cover the complete 5 × 5 physical resistance–inductance grid and both 310 A frequency branches at each of 23 feasible points. Fixed-frequency plant periodic orbits are locally contractive, while controller-in-the-loop tests settle on both branches within the tested horizon. The conclusions are local to the calibrated operating region and do not claim a global optimum or global nonlinear stability. The negative plant-orbit Floquet exponents reported here establish local contraction of the fixed-frequency physical periodic orbit; they are not interpreted as a global nonlinear closed-loop stability certificate. The normalized uncertainty square used for controller synthesis was constructed from rounded coefficient values. It includes conservative coefficient combinations that are not generated by a physical resistance–inductance pair, but the exact physical extreme of a₃₁ at L = 0.5L_nom slightly exceeds the rounded upper limit. Accordingly, the original square is treated as a convenient design parametrization rather than as a strict physical overbound. For verification, a residual-augmented coefficient box was reconstructed from the exact analytical extrema and the measured interpolation residuals; this coverage-preserving set retains the robust-performance conclusion. Full article

The opioid epidemic continues to be a major public health crisis in the United States, and nalmefene has recently emerged as an alternative opioid overdose reversal agent. However, little is known about pharmacists’ awareness and beliefs regarding nalmefene use and distribution. Therefore, the purpose of this study was to assess Alabama pharmacists’ knowledge, perceptions, and behavioral intentions related to nalmefene. This study used a cross-sectional survey design. Practicing pharmacists in the state of Alabama were eligible to participate, and were recruited through the Alabama Board of Pharmacy email listserv. An anonymous online survey was distributed via email, and respondents were eligible to enter a lottery for one of five $100 electronic gift cards. The survey instrument was developed by the investigators, adapted from prior overdose reversal literature and informed by the Theory of Planned Behavior. Primary outcome measures included: knowledge of nalmefene (12 items); perceived barriers regarding nalmefene stocking and recommendations (19 items); and Theory of Planned Behavior concepts including attitudes (15 items), subjective norms (perceived social support) (six items), perceived behavioral control (confidence in ability to stock or recommend nalmefene) (11 items), and behavioral intentions surrounding nalmefene stocking and recommendations (13 items). Outcomes were measured via multiple-choice (objective knowledge) and Likert-type scale (1 = strongly disagree, 5 = strongly agree) questions. Differences in mean scale scores across pharmacy setting (pharmacists employed in inpatient versus outpatient settings), community pharmacy type (chain vs. independent), and geographic location (rural vs. urban) were analyzed using two-sided Mann–Whitney U tests for non-parametric data or t-tests for parametric data, as appropriate. Multiple linear regression analyses examined associations between TPB constructs and pharmacists’ intentions to stock or recommend nalmefene over the next six months. All data were analyzed using SPSS statistical software version 29 with an alpha of 0.05. There were 119 pharmacist respondents (n = 25 inpatient, n = 94 outpatient). The majority were female (68.1%), White (94.1%), and worked in community settings (35.3% independent, 10.9% corporate, 10.1% big-box, 5.0% grocery). Median age was 42 years (IQR: 36–51). Awareness of nalmefene was low: only 39.5% had heard of it and 5.9% had seen it in practice. Inpatient versus outpatient pharmacists reported higher subjective norms external to the workplace (median [IQR]: 3.00 [2.33, 3.58] vs. 2.33 [1.67, 3.00]; p = 0.004) as well as subjective knowledge (mean [SD]: 2.96 [0.94] vs. 2.55 [0.74]; p = 0.023), while outpatient versus inpatient pharmacists reported higher perceived barriers related to fiscal/logistical concerns (mean [SD]: 3.45 [0.59] vs. 3.07 [0.66]; p = 0.009). Among community pharmacists, stocking and procurement confidence was higher among independent versus chain pharmacists (3.75 [3.00–4.17] vs. 3.17 [2.67–3.92]; p = 0.026). Furthermore, external workplace norms among urban pharmacists were higher than among rural pharmacists (2.50 [2.00–3.00] vs. 2.33 [1.67, 2.67]; p = 0.027), while stocking and procurement confidence was higher among rural than urban pharmacists (3.83 [2.83–4.50] vs. 3.17 [2.67–4.00]; p = 0.034). In adjusted reduced regression models, perceived behavioral control regarding nalmefene stocking and procurement (B = 0.189, 95% CI: 0.049, 0.329; p = 0.009), subjective norms external to the workplace (B = 0.232, 95% CI: 0.047, 0.418; p = 0.015), and interprofessional recommendation attitudes (B = 0.237, 95% CI: 0.037, 0.437; p = 0.021) positively predicted nalmefene implementation intentions. These findings suggest practice-setting differences that could be further explored for optimal implementation of nalmefene distribution and access, and highlight the need for targeted education, interprofessional collaboration, procurement guidance, and workflow tools to improve pharmacist preparedness for nalmefene services. Full article

Built-up expansion is reshaping landscapes across Nepal; however, consistent multi-temporal mapping remains challenging due to rugged terrain, fragmented settlements, and heterogeneous land-cover conditions. This study develops and evaluates a multi-product and terrain-informed workflow in Google Earth Engine for mapping built-up expansion across three physiographically contrasting districts of Nepal: Arghakhanchi, Lalitpur, and Chitwan, from 2017 to 2025. Annual predictor stacks were generated by integrating Sentinel-2 spectral bands and derived indices, Dynamic World built-up probabilities, and SRTM-derived elevation and slope variables. ESRI Global Land Cover datasets were used separately for auxiliary cross-product comparison and assessment of the mapped outputs. Preliminary yearly built-up masks were generated using district- and year-specific Random Forest classifications, followed by the post-classification constraints, and were subsequently integrated through cumulative expansion mapping. Accuracy assessment for 2017, 2021, and 2025 yielded overall accuracy values of 86.4–92.4%, built-up F1-scores of 84.7–91.3%, and Kappa coefficients of 0.81–0.91. Between 2017 and 2025, cumulative built-up extent expanded by 8054.65 ha in Chitwan, 2406.20 ha in Arghakhanchi, and 2215.96 ha in Lalitpur; Arghakhanchi recorded the highest proportional increase (117.6%). The mapped expansion was comparatively dispersed in Arghakhanchi, concentrated within metropolitan and peri-urban areas in Lalitpur, and broader and corridor-oriented in Chitwan. Because previously detected built-up pixels were retained in subsequent cumulative outputs, the resulting extents were non-decreasing by construction and did not represent demolition or other land use reversals. Consequently, annual built-up expansion should not be interpreted as net annual land-cover change. The proposed framework provides a practical and transferable approach for comparative built-up expansion monitoring and urban growth assessment across contrasting physiographic settings. Full article

Ultrasound assessment of developmental dysplasia of the hip (DDH) in infants is highly operator-dependent, particularly during standard plane acquisition and Graf angle measurement. This paper proposes DDH-Net, a fully automated computer-aided assessment method for infant hip ultrasound that follows the complete Graf workflow. First, YOLOv8n-pose is used to detect eight anatomical structures and two ilium orientation keypoints for standard plane determination. Regions of interest (ROIs) are then cropped according to the detection results, and U-Net is employed to perform fine segmentation of the ilium, labrum, and lower limb of the ilium. Finally, Graf reference lines are constructed from the segmentation results to enable automatic measurement of the α and β angles. The study included 1409 infants, comprising 2648 ultrasound images and 50 ultrasound videos. On an independent test set of 529 images, the accuracy, sensitivity, and specificity of standard plane detection were 96.8%, 94.5%, and 100.0%, respectively. The model automatically saved 187 candidate frames from the videos, of which 93.1% were rated by experts as having no or only minor clinically relevant discrepancies; 48 of the 50 videos (96.0%) contained at least one clinically acceptable candidate frame. Compared with full-image segmentation, ROI-based segmentation reduced processing time by 33.7%. In 307 standard plane images, the mean absolute errors (MAE) for the α and β angles were 1.61° and 2.13°, respectively, with corresponding intraclass correlation coefficients (ICC) of 0.913 and 0.766. The complete pipeline achieved a processing speed of 28.87 frames per second, indicating that DDH-Net has the potential to provide efficient and interpretable analysis of infant hip ultrasound images. Full article

Accurate stability assessment of potentially unstable rock blocks is essential for the safe construction and operation of hydropower infrastructure. This study applies a comparative limit-equilibrium workflow to a single, well-characterized sliding-type rock block (156.7 m³) bounded by three discontinuities (J1 250°/35°, J2 305°/75°, J3 215°/80°) on a hydropower station slope; discontinuity attitudes were measured with a geological compass and a terrestrial three-dimensional laser scanner, and the slope surface was reconstructed by UAV photogrammetry. A common, fully documented parameter set is used by a conventional two-dimensional method, a block-dividing limit-equilibrium method, and a three-dimensional residual-thrust method with moment equilibrium. With the site-suggested shear strengths and the lower-bound cohesion as the representative value, the block-dividing method gives factors of safety of 1.169, 1.063 and 0.903 under natural, heavy-rainfall and seismic conditions at optimal azimuths of 265.8°, 266.3° and 269.1°; the corresponding two-dimensional values are 1.037, 0.904 and 0.729, and the residual-thrust values are 1.497, 1.324 and 1.049. The block-dividing factors are 12.7–23.9% above the two-dimensional profile, whereas the residual-thrust result lies a further 16–28% higher (44–47% above the two-dimensional profile); this over-estimate is traced to the steep (75°, 80°) lateral release planes and to a mesh- and lambda-sensitive column solution, and is therefore non-conservative. A cohesion sensitivity analysis with fixed friction angle shows that the factor varies by 51–57% across the suggested cohesion interval. Under code-specified targets of 1.30/1.20/1.05, horizontal anchorage requires 366/427/566 kN versus 1150/1470/1413 kN for a perpendicular-to-slope layout, so a horizontal scheme of about 0.6 MN is adopted. Finite-element validation and field piezometric/displacement monitoring data, unavailable for this block, are identified as required future work. Full article

Velocity distributions in curved open-channel flows exhibit strong three-dimensionality and nonlinear behavior, posing challenges to both accurate prediction and physical interpretation. Using measured velocity data from nine discharge–water-depth combinations in a laboratory 180° open-channel bend, this study developed an integrated eXtreme Gradient Boosting (XGBoost)–SHapley Additive exPlanations (SHAP) framework, with multiple linear regression (MLR), random forest (RF), and a back-propagation neural network (BPNN) used for comparison. Leave-one-condition-out cross-validation was used to evaluate the predictive accuracy and stability of the four models. A stratified sampling strategy was then adopted to construct the training dataset, allowing information from all flow regimes to contribute to robust parameter calibration; the two data-partitioning strategies yielded broadly comparable predictive performance. Using models trained with stratified sampling, multidimensional model evaluation was further conducted using global statistical metrics, segment-wise predictive performance, held-out extreme-condition tests, and measured–predicted agreement, among other criteria, with XGBoost consistently showing the best performance. Multilevel SHAP analyses quantified global feature importance, pairwise interactions, streamwise variations in feature contributions, SHAP–PDP dependence relationships, and condition-specific attribution. The SHAP results indicate a two-level attribution structure in the model: hydraulic variables jointly define the global velocity baseline, and their contribution signs can switch between positive and negative. Spatial variables characterize cross-sectional velocity redistribution. Strong discharge–depth interaction is associated with width-to-depth-ratio-dependent adjustment of the bend flow field. The proposed framework establishes a complete experiment-driven prediction–mechanism interpretation workflow for sharply curved open-channel flow and provides new quantitative insight into model-represented multifactor hydrodynamic interactions in open-channel bends. Full article