Search Results (610)

This study investigated the effects of incorporating rice husk ash (RHA) as a partial replacement for cement on the properties of concrete. To determine the optimal replacement level, RHA was used to replace cement in varying proportions, ranging from 0% to 25% in 5% increments. The mix with 0% RHA served as the control. The properties evaluated included setting time, density, and compressive strength. The results revealed that blending RHA with cement increased the initial setting time. This was attributed to the lower calcium oxide (CaO₂) content of RHA, which slows early-age hydration reactions. Conversely, the final setting time was reduced due to the pozzolanic activity of RHA, which enhances later-stage reactions. Additionally, the inclusion of RHA resulted in a decrease in concrete density, owing to its lower specific gravity and bulk density compared to Portland cement. Despite this, RHA-modified specimens exhibited higher compressive strengths than the control specimens. This strength enhancement was linked to the formation of additional calcium–silicate–hydrate (C-S-H) gel due to the pozzolanic reaction between amorphous silica in RHA and calcium hydroxide (CaOH) from hydration reaction. The gel fills concrete voids at the microstructural level, producing a denser and more compact concrete matrix. Based on the balance between strength and durability, the optimal RHA replacement level was identified as 10%. Full article

The chemical composition, mineral composition, and mineral distribution characteristics of steel slag were characterized through petrographic analysis, X-ray diffraction (XRD), and particle size analysis. Limestone, silica, and silicomanganese slag were blended with converter steel slag to fabricate a reconstructed steel slag. Through burden calculation, the chemical composition ratio of this reconstructed steel slag approximated the silicate phase region. The high-temperature reconstruction process outside the furnace was simulated through reheating. The composition, structure, and cementitious characteristics of the reconstructed steel slag were investigated through X-ray diffraction (XRD), FactSage software (FactSage version 7.0 (GTT-Technologies, Aachen, Germany, 2015))analysis, scanning electron microscopy–energy dispersive spectroscopy (SEM–EDS) analysis, setting time determination, compressive strength measurement, and thermodynamic computation. The findings indicated that the primary mineral compositions of the reconstructed steel slag were predominantly silicates, such as Ca₃Al₂O₆, Ca₂SiO₄, Ca₂MgSi₂O₇, Ca₂Al(AlSiO₇), Ca₂(SiO₄), and FeAlMgO₄. In comparison with the original steel slag, these compositions underwent substantial alterations. The α′-C₂S phase appears at 1100 K and gradually transforms into α-C₂S at 1650 K. The liquid phase begins to precipitate at approximately 1550 K. Spinel exists in the temperature range from 1300 to 1700 K, and Ca₃MgSi₂O₈ melts into the liquid phase at 1400 K. As the temperature increases to 1600 K, the minerals C₂AF, Ca₂Fe₂O₅, and Ca₂Al₂O₅ gradually melt into the liquid phase. Melilite melts into the liquid phase at 1700 K. It was observed that the initial and final setting times of the reconstructed steel slag exhibited reductions of 7 and 43 min, respectively, in comparison to those of the original steel slag. In comparison with steel slag, the compressive strength of the reconstructed steel slag exhibited an increase of 0.6 MPa at the 3-day strength stage, 1.6 MPa at the 7-day strength stage, and 3.4 MPa at the 28-day strength stage. The reduction in setting time and the enhancement in compressive strength verified the improved cementitious activity of the reconstructed steel slag. Thermodynamic calculations of the principal reactions of the reconstructed steel slag at elevated temperatures verified that the primary reaction at 1748 K is thermodynamically favorable. Full article

Intelligent recommender systems are essential for identifying at-risk students and personalizing learning through tailored resources. Accurate prediction of student performance enables these systems to deliver timely interventions and data-driven support. This paper presents the application of machine learning models to predict final exam grades in a university-level programming course, leveraging multi-modal student data to improve prediction accuracy. In particular, a recent raw dataset of students enrolled in a programming course across 36 class sections from the Fall 2024 and Winter 2025 terms was initially processed. The data was collected up to one month before the final exam. From this data, a comprehensive set of features was engineered, including the student’s background, assessment grades and completion times, digital learning interactions, and engagement metrics. Building on this feature set, six machine learning prediction models were initially developed using data from the Fall 2024 term. Both training and testing were conducted on this dataset using cross-validation combined with hyperparameter tuning. The XGBoost model demonstrated strong performance, achieving an accuracy exceeding 91%. To assess the generalizability of the considered models, all models were retrained on the complete Fall 2024 dataset. They were then evaluated on an independent dataset from Winter 2025, with XGBoost achieving the highest accuracy, exceeding 84%. Feature importance analysis has revealed that the midterm grade and the average completion duration of lab assessments are the most influential predictors. This data-driven approach empowers instructors to proactively identify and support at-risk students, enabling adaptive learning environments that deliver personalized learning and timely interventions. Full article

Blockchain technology has attracted considerable attention in the supply chain literature for its potential to enhance operational traceability, transparency, and trust, as well as to advance greening initiatives. Given current supply chain configurations, exploring barriers to implementation is a consequential agenda, and current studies have devoted substantial effort to identifying and offering guidance to address them. Despite recent findings, insights into how blockchain technology adoption can support green supply chain management are missing, particularly in the maritime sector, which receives limited attention. Thus, this work outlines a methodological approach to examine the suitability of maritime routes for addressing barriers to implementing blockchain technology in green supply chain management. Viewing the evaluation as a multi-criteria decision-making (MCDM) problem, the proposed approach performs the following actions on a case study evaluating four maritime lines. Firstly, from the 13 identified barriers in the literature review and expert interviews, nine relevant barriers were determined after one round of a Delphi process. These barriers eventually comprise the set of evaluation criteria. Secondly, to satisfy the assumption of criterion independence in most MCDM methods, this work proposes a novel extended grey relational analysis (GRA) that allows for the measurement of criterion independence based on the concept of grey relational space. Proposed here for the first time, the extended GRA offers a distribution-free overall independence index for each criterion based on pattern similarity. Finally, an integration of SWARA (Stepwise Weight Assessment Ratio Analysis) and ARAS (Additive Ratio Assessment) methods under Z-information is developed to address the evaluation problem involving expert judgments in a highly uncertain decision-making context. Results show that transaction-level uncertainty is the most critical barrier to blockchain adoption, followed by technology risks and higher sustainability costs. Among the four maritime lines, Line 3 is best prepared for a blockchain-enabled green supply chain. The agreement between these results and those of other MCDM methods is shown in the comparative analysis. Also, ranking remains unchanged even when the criteria weights are adjusted. The proposed approach provides a computationally efficient and tractable framework for maritime managers to make informed decisions about blockchain adoption to promote green supply chains. Full article

Semantic segmentation of UAV–acquired RGB orthomosaics is a key component for quantifying vegetation cover and monitoring phenology in precision agriculture. This study evaluates a representative set of CNN–based architectures (U–Net, U–Net Xception–Style, SegNet, DeepLabV3+) and Transformer–based models (Swin–UNet/Swin–Transformer, SegFormer, and Mask2Former) under a unified and reproducible protocol. We propose a transfer–and–consolidation workflow whose performance is assessed not only through region–overlap and pixel–wise discrepancy metrics, but also via boundary–sensitive criteria that are explicitly linked to orthomosaic–scale vegetation–cover estimation by pixel counting under GSD (Ground Sample Distance) control. The experimental design considers a transfer scenario between morphologically related crops: initial training on Opuntia spp. (prickly pear), direct (“zero–shot”) inference on Agave salmiana, fine–tuning using only 6.84% of the agave tessellated set as limited target–domain supervision, and a subsequent consolidation stage to obtain a multi–species model. The evaluation integrates IoU, Dice, RMSE, pixel accuracy, and computational cost (time per image), and additionally reports the BF score and HD95 to characterize contour fidelity, which is critical when area is derived from orthomosaic–scale masks. Results show that Transformer-based approaches tend to provide higher stability and improved boundary delineation on Opuntia spp., whereas transfer to Agave salmiana exhibits selective degradation that is mitigated through low–annotation–cost fine-tuning. On Opuntia spp., Mask2Former achieves the best test performance (IoU 0.897 +/− 0.094; RMSE 0.146 +/− 0.002) and, after consolidation, sustains the highest overlap on both crops (IoU 0.894 +/− 0.004 on Opuntia and IoU 0.760 +/− 0.046 on Agave), while preserving high contour fidelity (BF score 0.962 +/− 0.102/0.877 +/− 0.153; HD95 2.189 +/− 3.447 px/8.458 +/− 16.667 px for Opuntia/Agave), supporting its use for final vegetation–cover quantification. Overall, the study provides practical guidelines for architecture selection under hardware constraints, a reproducible transfer protocol, and an orthomosaic–oriented implementation that facilitates integration into agronomic and remote–sensing workflows. Full article

In this paper, we take the longitudinal two-stage and two-yard EMU (Electric Multiple Unit) depot as an example and discusses the optimization challenges of the first-level maintenance shunting operation plan under the background of limited maintenance capacity. A multi-objective programming is constructed, which adopts the lexicographic ordering method and aims to minimize the occupancy time of key line areas and the number of train storage times. In order to enhance the flexibility and solution efficiency of the shunting operation plan, we design an efficient three-stage strategy algorithm. Specifically, in the first stage, the genetic and mutation rules are integrated, and the fast iterative advantage of the genetic algorithm is utilized to solve the time decision variables in the optimization problem. In the second stage, the allocation of track occupancy variables is further solved. The third stage focuses on the optimized allocation of maintenance team variables to ensure the scientific scheduling of maintenance resources. Finally, a validation experiment was conducted using the maintenance tasks of 19 EMU sets as the test scenario. The results indicate that when the number of maintenance teams is set to 4, an optimal balance between maintenance efficiency and operational cost is achieved, the occupancy duration of key line zones reaches 3034 min (the theoretical optimum), the number of maintenance teams is reduced by 33.33% compared to the initial 6 teams, and the number of storage operations is optimized to 27 times. Additionally, the algorithm’s solution time remains under 50 s, demonstrating significantly improved computational efficiency. Comparative experiments with baseline algorithms show that the proposed method reduces the occupancy duration of key line zones by up to 0.49%, decreases the number of storage operations by 14 times, and advances the maximum completion time by 20 min. In summary, the proposed method provides solid theoretical support for the formulation of maintenance plans and shunting schedules in EMU depots. Particularly in complex scenarios with limited maintenance capacity, it offers innovative and robust decision-making foundations, demonstrating significant practical guidance value. Full article

Background and Objectives: Keratoconus is a progressive corneal ectatic disorder and a leading cause of corneal transplantation in developed countries. Early detection is critical for initiating timely interventions such as corneal cross-linking, which can halt disease progression and preserve long-term visual function. This review aims to synthesize current diagnostic approaches for early keratoconus detection and assess their clinical impact on visual outcomes. Materials and Methods: A comprehensive literature search was conducted across PubMed/MEDLINE, Web of Science, Google Scholar, Scopus and the Cochrane Library through September 2025. Search terms included “early keratoconus,” “subclinical keratoconus,” “forme fruste keratoconus,” “keratoconus detection,” “corneal topography,” “corneal tomography,” “anterior segment optical coherence tomography (AS-OCT),” “corneal biomechanics,” “artificial intelligence,” “genetic risk, “environmental factors”, and “machine learning.” Two independent reviewers analyzed the data. Studies were included if they investigated diagnostic modalities for early-stage keratoconus and discussed their relevance to visual outcomes. Results: One hundred and seven studies were included in the final review. Four diagnostic modalities demonstrated consistent clinical value: 1. corneal topography for assessing anterior surface irregularities; 2. corneal tomography, currently regarded as the gold standard due to its ability to detect early posterior elevation and pachymetric changes; 3. AS-OCT for epithelial and stromal profiling; and 4. biomechanical assessments, which evaluate corneal tissue stability prior to structural alterations. Artificial intelligence, when integrated with imaging data, enhances diagnostic sensitivity and standardizes interpretation across clinical settings. Conclusions: Early keratoconus detection is crucial for preserving vision; and integrating multimodal, AI-supported diagnostics into routine care—especially for high-risk groups—enhances accuracy, improves outcomes, and reduces progression rates of disease. Full article

Winter jujube harvesting is traditionally labor-intensive, yet declining labor availability and rising costs necessitate robotic automation to maintain agricultural competitiveness. Path planning for robotic arms in orchards faces challenges due to the unstructured, dynamic environment containing densely packed fruits and branches. To overcome the limitations of existing robotic path planning methods, this research proposes BMGA-RRT Connect (BVH-based Multilevel-step Gradient-descent Adaptive RRT), a novel algorithm integrating adaptive multilevel step-sizing, hierarchical Bounding Volume Hierarchy (BVH)-based collision detection, and gradient-descent path smoothing. Initially, an adaptive step-size strategy dynamically adjusts node expansions, optimizing efficiency and avoiding collisions; subsequently, a hierarchical BVH improves collision-detection speed, significantly reducing computational time; finally, gradient-descent smoothing enhances trajectory continuity and path quality. Comprehensive 2D and 3D simulation experiments, dynamic obstacle validations, and real-world winter jujube harvesting trials were conducted to assess algorithm performance. Results showed that BMGA-RRT Connect significantly reduced average computation time to 2.23 s (2D) and 7.12 s (3D), outperforming traditional algorithms in path quality, stability, and robustness. Specifically, BMGA-RRT Connect achieved 100% path planning success and 90% execution success in robotic harvesting tests. These findings demonstrate that BMGA-RRT Connect provides an efficient, stable, and reliable solution for robotic harvesting in complex, unstructured agricultural settings, offering substantial promise for practical deployment in precision agriculture. Full article

Vegetation concrete is a composite material integrating plant growth and concrete technology. In this study, solid waste materials (phosphogypsum and recycled aggregates) were utilized to prepare vegetation concrete. Semi-hydrated phosphogypsum (HPG) was used to replace ordinary Portland cement as a cementitious material in a gradient manner, while recycled coarse aggregates (RCAs) fully replaced natural crushed stone. The basic properties of phosphogypsum–recycled aggregate-based vegetation concrete were analyzed, and X-ray diffraction (XRD) and scanning electron microscopy (SEM) were employed to characterize the hydration products of vegetation concrete with different mix ratios. The results indicated that replacing cement with HPG exerted a significant alkali-reducing effect and provided favorable cementitious strength. When the porosity was 24% and the HPG content was 50%, the vegetation concrete exhibited optimal performance: the 28-day compressive strength reached 12.3 MPa, and the pH value was 9.7. Recycled aggregates had a minimal impact on strength. When 0.5% sodium gluconate was added as a retarder, the initial setting time was 97 min and the final setting time was 192 min, which met construction requirements with little influence on later-stage strength. Microscopic analysis revealed that the early strength (3d–7d) of vegetation concrete was primarily contributed by CaSO₄·2H₂O crystals (the hydration product of HPG), while the later-stage strength was supplemented by C-S-H (the hydration product of cement). Planting tests showed that Tall Fescue formed a lawn within 30 days; at 60 days, the plant height was 18 cm and the root length was 6–8 cm. Some roots grew along the sidewalls of concrete pores and penetrated the 5 cm thick vegetation concrete slab, demonstrating good growth status. Full article

Expansive soil is prone to significant swelling and shrinkage deformation with changes in moisture conditions, posing serious safety hazards to engineering construction. This study focuses on alkali-activated self-compacting fluid-solidified soil (ASFS) and systematically explores the regulatory effect of expansive soil with different dosages (0–100%) on its properties. This study analyzes the influence of expansive soil on the setting time, hydration characteristics, autogenous shrinkage, and compressive strength of ASFS while verifying the feasibility of this method for solidifying expansive soil through microstructural analysis. The results show that, with the increase in content of expansive soil, the initial and final setting times of ASFS were prolonged by 0.08–1.58 times and 0.08–1.29 times, respectively. Although expansive soil inhibited the hydration of ASFS, it could compensate for autogenous shrinkage through the expansion effect of clay minerals, reducing the autogenous shrinkage by 13.4–51.2%. Furthermore, the optimal dosage of expansive soil in ASFS is 60%. Compared with the control group, the 7d compressive strength of ASFS increases by 52.4%, the strength after 3d water immersion rises by 62.6%, and the strength after eight wet–dry cycles still remains 10% higher. This optimal dosage achieves the best balance between mechanical properties, water stability, and shrinkage resistance of ASFS, providing a reliable technical reference for the efficient utilization of expansive soil in engineering. Full article

The strong uncertainty of renewable energy poses significant reliability and safety challenges for the coordinated operation of multi-integrated energy systems (MIES). To address this, a data-driven two-stage distributed robust collaborative optimization scheduling model for MIES is proposed, based on a spatiotemporal fusion conditional diffusion model (STF-CDM). First, to more accurately capture the uncertainty in renewable energy output, the model utilizes a scenario set generated by the STF-CDM model and reduced via the K-means clustering algorithm as the initial renewable energy scenarios for the distributed robust optimization set. The STF-CDM model employs a Temporal module component (TMC) unit composed of Transformer time-series modules and a Spatial module component (SMC) unit composed of CNN neural networks for feature extraction and fusion of time-series and spatial-series data. Second, a benefit allocation method based on multi-energy trading contribution rates is proposed to achieve equitable distribution of cooperative gains. Finally, to protect participant privacy and enhance computational efficiency, an alternating direction multiplier method (ADMM) coupled with parallelizable column and constraint generation (C&CG) is employed to solve the energy trading problem. The case analysis results demonstrate that the STF-CDM model proposed in this study exhibits superior performance in addressing the uncertainty of renewable energy output. Concurrently, the asymmetric Nash game mechanism and the ADMM-C&CG solution algorithm proposed in this study achieve a fair and reasonable distribution of benefits among all participants when handling energy transactions and cooperative gains. This is accomplished while ensuring system robustness, economic efficiency, and privacy. Full article

Ensuring optimal food hygiene is essential for food safety and preventing foodborne illness, although the importance of food hygiene is often overlooked in the household kitchen setting. Adequate, good hygiene practices in the domestic environment are equally important as their implementation in any other food preparation environment, like in the food industry. The current review encompasses research data on the prevalence and isolation of major foodborne pathogenic bacteria (Campylobacter, Salmonella, Listeria monocytogenes, Staphylococcus aureus, Escherichia coli pathotypes, and Clostridium perfringens) from household kitchen equipment, as well as food cleaning utensils used in the kitchen, such as sponges, brushes, dishcloths, and hand towels. The most common bacterial pathogen present in the domestic environment is S. aureus. The latter can be transmitted orally, either via direct hand contact with contaminated kitchen surfaces and/or cleaning utensils, or indirectly through the consumption of contaminated food due to cross-contamination during food preparation (e.g., portioning prepared meat on the same cutting board surface and with the same knife previously used to cut fresh leafy vegetables). Moreover, research findings on the hygiene of food cleaning utensils demonstrate that (i) sponges have the highest microbial load compared to all other cleaning utensils, (ii) brushes are less contaminated and more hygienic than sponges, thus safer for cleaning cutlery and kitchen utensils, and (iii) kitchen dishcloths and hand towels positively contribute to cross-contamination since they are frequently used for multiple purposes at the same time (e.g., drying hands and wiping/removing excess moisture from dishes). Finally, the present review clearly addresses the emerging issue of antimicrobial resistance (AMR) in bacterial pathogens and the role of the domestic kitchen environment in AMR dissemination. These issues add complexity to foodborne risk management, linking household practices to broader AMR stewardship initiatives. Full article

Energy-efficient Train Control (EETC) strategy needs to meet safety, punctuality, and energy-saving requirements during train operation, and puts forward higher requirements for online use and adaptive ability. In order to meet the above requirements and reduce the dependence on an accurate mathematical model of train operation, this paper proposes a train-speed trajectory-optimization method combining data-driven energy consumption estimation and deep reinforcement learning. First of all, using real subway operation data, the key unit basic resistance coefficient in train operation is analyzed by regression. Then, based on the identified model, the energy consumption experiment data of train operation is generated, into which Gaussian noise is introduced to simulate real-world sensor measurement errors and environmental uncertainties. The energy consumption estimation model based on a Backpropagation (BP) neural network is constructed and trained. Finally, the energy consumption estimation model serves as a component within the Deep Deterministic Policy Gradient (DDPG) algorithm environment, and the action adjustment mechanism and reward are designed by integrating the expert experience to complete the optimization training of the strategy network. Experimental results demonstrate that the proposed method reduces energy consumption by approximately 4.4% compared to actual manual operation data. Furthermore, it achieves a solution deviation of less than 0.3% compared to the theoretical optimal baseline (Dynamic Programming), proving its ability to approximate global optimality. In addition, the proposed algorithm can adapt to the changes in train mass, initial set running time, and halfway running time while ensuring convergence performance and trajectory energy saving during online use. Full article

Accurate recognition of complex human activities from wearable sensors plays a critical role in sports analytics and human performance monitoring. However, the high dimensionality and redundancy of raw inertial data can hinder model performance and interpretability. This study proposes a hybrid feature selection framework that combines Minimum Redundancy Maximum Relevance (MRMR) and Regularized Neighborhood Component Analysis (RNCA) to improve classification accuracy while reducing computational complexity. Multi-sensor inertial data were collected from field hockey players performing six activity types. Time- and frequency-domain features were extracted from four body-mounted inertial measurement units (IMUs), resulting in 432 initial features. MRMR, combined with Pearson correlation filtering (|ρ| > 0.7), eliminated redundant features, and RNCA further refined the subset by learning supervised feature weights. The final model achieved a test accuracy of 92.82% and F1-score of 86.91% using only 83 features, surpassing the MRMR-only configuration and slightly outperforming the full feature set. This performance was supported by reduced training time, improved confusion matrix profiles, and enhanced class separability in PCA and t-SNE visualizations. These results demonstrate the effectiveness of the proposed two-stage feature selection method in optimizing classification performance while enhancing model efficiency and interpretability for real-time human activity recognition systems. Full article

Geopolymers, as sustainable alternatives to traditional cementitious materials, offer superior mechanical and durability properties; however, they face challenges with rapid setting, particularly in fly ash–slag systems. Retarders play a crucial role in tailoring the setting behavior and workability of geopolymers, especially in applications where extended setting time or placement under challenging conditions is required. Geopolymers, unlike traditional Portland cement, undergo a rapid alkali-activation process involving dissolution, polymerization, and hardening of aluminosilicate materials. This can lead to very short setting times, particularly at elevated temperatures. In this scenario, the present study investigates the effect of different retarders-including cellulose, starch, borax, and their different combinations the setting time. The effectiveness of a retarder depends on the geopolymer formulation, including the type of precursor, activator, and curing conditions. The initial and final setting times improved by the addition of retarders, whereas most of the retarders had a negative effect on compressive strength. The optimum retarder combination was starch and borax, with a remarkable improvement in setting time and a positive result on the compressive strength, while maintaining reasonable workability. The retarder was equally effective under both ambient and oven-cured conditions and for different mix proportions of fly ash (FA) and slag, indicating that its effectiveness depends only on the type of precursors used. The study reveals the use of borax along with cellulose- or sugar-based compounds, which balances the reaction kinetics, resulting in balanced mechanical characteristics. Full article