Search Results (7,150)

The development of functional biomaterials based on natural polymers has gained increasing relevance due to the growing demand for sustainable and bioactive alternatives for biomedical and technological applications. In this study, chitosan was obtained from shrimp exoskeletons and used to formulate active films enriched with Mexican propolis, aiming to evaluate the influence of the extract on the physicochemical and functional properties of the resulting biomaterial. Propolis was incorporated into the chitosan film-forming solution at a final concentration of 1.0% (v/v). The propolis employed met the requirements of the Mexican Official Standard NOM-003-SAG/GAN-2017 regarding flavonoid content, total phenolic compounds, and antimicrobial activity; additionally, it was evaluated through antioxidant activity, hemolysis, and acute toxicity (LD₅₀) assays to provide a broader biological and safety assessment. The extracted chitosan exhibited a degree of deacetylation of 74% and characteristic FTIR spectral features comparable to those of commercial chitosan, confirming the quality of the obtained polymer. Chitosan–propolis films exhibited antimicrobial activity against Staphylococcus aureus, Escherichia coli, and Candida albicans, whereas pure chitosan films showed no inhibitory effect. Thermal analyses (TGA/DSC) revealed a slight reduction in thermal stability due to the incorporation of thermolabile polyphenolic compounds, along with increased thermal complexity of the system. SEM observations demonstrated reduced microbial adhesion and marked morphological damage in microorganisms exposed to the functionalized films. Overall, the incorporation of Mexican propolis enabled the development of a hybrid biomaterial with enhanced antimicrobial performance and potential application in wound dressings and bioactive coatings. Full article

Artificial neural networks (ANNs) are widely used in engineering prediction, but excessive input dimensionality can reduce both accuracy and efficiency. This study proposes a two-stage feature-reduction framework, Feature Importance Ranking and Redundancy Elimination (FIRRE), to optimize ANN inputs by removing weakly informative and redundant variables. In Stage 1, four complementary ranking methods, namely Pearson correlation, recursive feature elimination, random forest importance, and F-test scoring, are combined into an ensemble importance score. In Stage 2, highly collinear features (ρ > 0.95) are pruned while retaining the more informative variable in each pair. FIRRE is evaluated on 32 civil engineering datasets spanning materials, structural, and environmental applications, and benchmarked against Principal Component Analysis, variance-threshold filtering, random feature selection, and K-means clustering. Across the benchmark suite, FIRRE consistently achieves competitive or improved predictive performance while reducing input dimensionality by 40% on average and decreasing computation time by 10–60%. A dynamic modulus case study further demonstrates its practical value, improving R² from 0.926 to 0.966 while reducing inputs from 25 to 7. Overall, FIRRE provides a practical, robust framework for simplifying ANN inputs and improving efficiency in civil engineering prediction tasks. Full article

Manual window cleaning in high-rise urban buildings is labor-intensive, risky, and resource-inefficient. This study addresses these challenges by investigating a resource-aware mechatronic architecture through the design, development, and experimental validation of a modular Automated Window Cleaning System (AWCS). Unlike conventional open-loop solutions, the AWCS integrates mechanical scrubbing with a closed-loop fluid management system, featuring precise dispensing and vacuum-assisted recovery. The system is governed by a deterministic finite state machine implemented on an ESP32 microcontroller, enabling low-latency IoT connectivity and autonomous operation. Two implementation variants—integrated and retrofit—were validated to ensure structural adaptability. Experimental results across 30 cycles demonstrate a cleaning efficiency of ~2 min/m², a water consumption of <150 mL/m² (representing a >95% reduction compared to manual methods), and an optical cleaning efficacy of 96.9% ± 1.4%. Safety protocols were substantiated through a calculated mechanical safety factor of 6.12 for retrofit applications. This research establishes the AWCS as a sustainable, safe, and scalable solution for autonomous building maintenance, contributing to the advancement of resource-circular domestic robotics and smart home automation. Full article

Controlling bottom hole pressure (BHP) during tripping-out is a key challenge in ultra-deep well drilling. Under high-temperature and high-pressure (HTHP) conditions, ultra-deep wells feature long tripping-out cycles, where thermal effects are prone to causing BHP reduction and increasing kick risk. However, existing pressure control technologies struggle to adapt to the requirements of narrow safe density windows in deep formations. This study establishes a transient tripping-out temperature field model, taking the PS6 ultra-deep vertical well as a case study to calculate the variations in temperature, equivalent static density (ESD), and BHP during tripping-out at 2910 m and 9026 m. A weighted drilling fluid supplementation method is presented, with supplementary parameters designed and its feasibility verified. The results indicate that during the entire tripping-out process, the bottom hole temperature at 2910 m increases by 17.5 °C and BHP rises by 0.016 MPa; at 9026 m, the temperature increases by 72.6 °C and BHP decreases by 2.410 MPa. Compared with the traditional “heavy mud cap” technology, the presented method can control BHP within a smaller fluctuation range (within 0.339 MPa) during tripping-out, better adapting to the safe tripping requirements of narrow safe density windows in deep formations and effectively mitigating kick risk. Full article

The causes of agricultural drought are complex, and its actual occurrence process is often characterized by rapid onset in terms of time and small scale in terms of space. Monitoring agricultural drought using satellite remote sensing with low spatial resolution makes it difficult to accurately capture the details of small-scale drought events. High-resolution satellite remote sensing has relatively long revisit cycles, making it difficult to capture the rapid evolution of drought conditions. Furthermore, the occurrence of agricultural drought is linked to multiple factors including precipitation, evapotranspiration, soil properties, and crop physiological characteristics. Consequently, relying on a single variable or indicator is insufficient for multidimensional monitoring of agricultural drought. This study takes Hebi City, Henan Province as the research area. It uses Sentinel-1 satellite data (HV, VV), Sentinel-2 data (NDVI, B2, B11), elevation, slope, aspect, and GPM precipitation data from 2019 to 2024 as independent variables. Three machine learning algorithms—Random Forest (RF), Random Forest-Recursive Feature Elimination (RF-RFE), and eXtreme Gradient Boosting (XGBoost)—were employed to construct a multi-dimensional agricultural drought monitoring model at the field scale. Additionally, the study verified the sensitivity of different environmental variables to agricultural drought monitoring and analyzed the accuracy performance of different machine learning algorithms in agricultural drought monitoring. The research results indicate that under the condition of full-factor input, all three models exhibit the optimal predictive performance. Among them, the XGBoost model performs the best, with the smallest Relative Root Mean Square Error (RRMSE) of 0.45 and the highest Correlation Coefficient (R) of 0.79. The absence of Digital Elevation Model (DEM) data impairs the models’ ability to capture the patterns of key features, which in turn leads to a reduction in predictive accuracy. Meanwhile, there is a significant correlation between model performance and sample size. Ultimately, the constructed XGBoost model takes the lead with an accuracy of 89%, while the accuracies of Random Forest (RF) and Random Forest-Recursive Feature Elimination (RF-RFE) are 88% and 86%, respectively. Based on these three drought monitoring models, this study further monitored a drought event that occurred in Hebi City in 2023, presented the spatiotemporal distribution of agricultural drought in Hebi City, and applied the Mann–Kendall test for time series analysis, aiming to identify the abrupt change process of agricultural drought. Meanwhile, on the basis of the research results, the feasibility of verifying drought occurrence using irrigation signals was discussed, and the potential reasons for the significantly lower drought occurrence probability in the western mountainous areas of the study region were analyzed. Full article

To achieve accurate voltage sag prediction and early warning, thereby improving power quality, a hybrid voltage sag prediction framework is proposed by integrating Kernel Entropy Component Analysis (KECA) with an improved Northern Goshawk Optimization (NGO) algorithm for hyperparameter tuning of a CNN-LSTM model. First, to address the limitations of the original NGO, such as proneness to falling into local optima and high randomness of the initial population distribution, a refraction-opposition-based learning mechanism is introduced to enhance population diversity and expand the search space. Furthermore, a sine–cosine strategy (SCA) with nonlinear weight coefficients is integrated into the exploration phase to dynamically adjust the search step size, optimizing the balance between global exploration and local exploitation, thereby boosting convergence speed and accuracy. The improved algorithm (SCNGO) is then utilized to optimize the hyperparameters of the CNN-LSTM model. Second, KECA is applied to voltage-sag-related data to extract key features and eliminate redundant information, and the resulting dimensionally reduced data are fed as input to the SCNGO-CNN-LSTM model to further improve prediction performance. Experimental results demonstrate that the SCNGO-CNN-LSTM model outperforms other comparative models significantly across multiple evaluation metrics. Compared with NGO-CNN-LSTM, GWO-CNN-LSTM, and the original CNN-LSTM, the proposed method achieves a mean squared error (MSE) reduction of 53.45%, 44.68%, and 66.76%, respectively. The corresponding root mean squared error (RMSE) is decreased by 25.33%, 18.61%, and 36.92%, while the mean absolute error (MAE) is reduced by 81.23%, 77.04%, and 86.06%, respectively. These results confirm that the proposed framework exhibits superior feature representation capability and significantly improves voltage sag prediction accuracy. Full article

Secchi disk depth (SDD) is a widely critical indicator of water transparency. However, existing retrieval models often suffer from limited transferability and biased predictions when applied to optically diverse waters. Here, we compiled a dataset of 6218 paired in situ SDD and remote sensing reflectance (R_rs) measurements to evaluate model generalization. We benchmarked nine machine learning (ML) models (RF, KNN, SVM, XGB, LGBM, CAT, RealMLP, BNN-MCD, and MDN) under three validation scenarios with progressively decreasing training-test overlap: Random, Waterbody, and Cross-Optical Water Type (Cross-OWT). Furthermore, SHAP analysis was employed to interpret feature contributions and relate model behaviors to optical properties. Results revealed a distinct scenario-dependent generalization gradient. Random splits yielded minimal bias. In contrast, Waterbody transfer consistently shifted predictions toward underestimation (SSPB: −16.9% to −3.8%). Notably, Cross-OWT extrapolation caused significant error inflation and a bias reversal toward overestimation (SSPB: 10.7% to 88.6%). Among all models, the Mixture Density Network (MDN) demonstrated superior robustness with the lowest overestimation (SSPB = 10.7%) under the Cross-OWT scenario. SHAP interpretation indicated that engineered indices, particularly NSMI, functioned as regime separators, with substantial shifts in feature attribution occurring at NSMI values between 0.4 and 0.6. Accordingly, feature sensitivity analysis showed that removing band ratios and indices improved Cross-OWT robustness for several classical ML models. For instance, KNN exhibited a significant reduction in Median Symmetric Accuracy (MdSA) from 96% to 40% after feature reduction. These findings highlight that model applicability must be evaluated under scenario-specific conditions, and feature engineering strategies require rigorous testing against optical regime shifts to ensure generalization. Full article

Endometrial lesions are common in gynecology, exhibiting considerable clinical heterogeneity across different subtypes. Although ultrasound imaging is the preferred diagnostic modality due to its noninvasive, accessible, and cost-effective nature, its diagnostic performance remains highly operator-dependent, leading to subjectivity and inconsistent results. To address these limitations, this study proposes a hierarchical feature selection and optimization framework for endometrial lesions, aiming to enhance the objectivity and robustness of ultrasound-based diagnosis. Firstly, Kernel Principal Component Analysis (KPCA) is employed for nonlinear dimensionality reduction, retaining the top 1000 principal components. Secondly, an ensemble of three filter-based methods—information gain, chi-square test, and symmetrical uncertainty—is integrated to rank and fuse features, followed by thresholding with Maximum Scatter Difference Linear Discriminant Analysis (MSDLDA) for preliminary feature selection. Finally, the Whale Migration Algorithm (WMA) is applied to population-based feature optimization and classifier training under the constraints of a Support Vector Machine (SVM) and a macro-averaged F1 score. Experimental results demonstrate that the proposed closed-loop pipeline of “kernel reduction—filter fusion—threshold pruning—intelligent optimization—robust classification” effectively balances nonlinear structure preservation, feature redundancy control, and model generalization, providing an interpretable, reproducible, and efficient solution for intelligent diagnosis in small- to medium-scale medical imaging datasets. Full article

Background: Autism Spectrum Disorder (ASD) is a type of neurodevelopmental disorder, and its exact causes are currently unknown. Neuroimaging research suggests that its clinical features are closely linked to alterations in brain functional network connectivity, yet the specific patterns and mechanisms underlying these abnormalities require further clarification. Methods: We recruited 36 children with ASD and 36 age- and sex-matched typically developing (TD) controls. Resting-state EEG data were used to construct static and dynamic low- and high-order functional networks across four frequency bands (δ, θ, α, β). Graph-theoretical metrics (clustering coefficient, characteristic path length, global efficiency, local efficiency) and state entropy were applied to characterize network topology and dynamic transitions between integration and segregation. Additionally, between-frequency networks were built for six band pairs (δ-θ, δ-α, δ-β, θ-α, θ-β, α-β), and network global measures quantified cross-frequency interactions. Results: Low-order networks in ASD showed increased δ and β connectivity but decreased θ and α connectivity. High-order networks demonstrated increased δ connectivity, reduced α connectivity, and mixed alterations in θ and β. Graph-theoretical analysis revealed pronounced α-band topological disruptions in ASD, reflected by a lower clustering coefficient and efficiency and higher characteristic path length in both low- and high-order networks. Dynamic analysis showed no significant entropy changes in low-order networks, while high-order networks exhibited time- and frequency-specific abnormalities, particularly in δ and α (0.5 s window) and δ (6 s window). Between-frequency analysis showed enhanced β-related coupling in low-order networks but widespread reductions across all band pairs in high-order networks. Conclusions: Young children with ASD exhibit coexisting hypo- and hyper-connectivity, disrupted network topology, and abnormal temporal dynamics. Integrating hierarchical, dynamic, and cross-frequency analyses offers new insights into ASD neurophysiology and potential biomarkers. Full article

Canine ehrlichiosis, caused by Ehrlichia canis, represents a relevant challenge in veterinary medicine, particularly in resource-limited settings where access to laboratory-based diagnostics may be constrained. This pilot and exploratory study aimed to evaluate the feasibility of a portable electronic olfactometer as a non-invasive screening approach, based on the analysis of volatile organic compounds (VOCs) present in breath, saliva, and hair samples from dogs. Signals were acquired using an array of eight metal-oxide (MOX) gas sensors (MQ and TGS series). After preprocessing, principal component analysis (PCA) was applied for dimensionality reduction, and the resulting features were analyzed using supervised machine-learning classifiers, including AdaBoost, support vector machines (SVM), k-nearest neighbors (k-NN), and Random Forests (RF). A total of 38 dogs (19 PCR-confirmed infected cases and 19 controls) were analyzed, generating 114 samples evenly distributed across the three biological matrices. Among the evaluated models, SVM showed the most consistent performance, particularly for saliva samples, achieving an accuracy, sensitivity, and precision of 94.7% (AUC = 0.964). In contrast, breath and hair samples showed lower discriminative performance. Given the limited sample size and the exploratory nature of the study, these results should be interpreted as preliminary; nevertheless, they suggest that electronic olfactometry may represent a complementary, low-cost, non-invasive screening tool for future research on canine ehrlichiosis, rather than a standalone diagnostic method. Full article

Surface water is essential for sustaining ecosystems and supporting human socio-economic development, yet pollution from urbanization increasingly threatens its ecological sustainability. The accurate prediction of dissolved oxygen (DO), as an important indicator of water quality, is crucial for water resource protection. To address the methodological gaps in current research, we propose a hybrid deep learning model (GCG) that integrates spatiotemporal correlations to enhance DO prediction accuracy through the systematic exploitation of latent data dependencies. This study proposes a three-stage modeling framework: (1) A novel adjacency matrix construction methodology based on Pearson correlation coefficients is developed to quantify spatial correlations between monitoring stations, enabling spatial feature aggregation via graph convolutional networks (GCNs); (2) the spatially enhanced features are subsequently processed through 1D convolutional neural networks (CNNs) to capture temporal local patterns; (3) model performance is comprehensively evaluated using four metrics: R², RMSE, MAE, and MAPE. The proposed model was implemented for DO prediction in Lake Taihu, China. Experimental results demonstrate that compared to conventional adjacency matrix construction methods, the Pearson correlation-based adjacency matrix confers advantages, achieving at least a 5% reduction in RMSE and over 10% improvement in MAE and MAPE. Furthermore, the GCG model outperformed the comparison model, with an R² enhancement of 8%, while reducing RMSE and MAE by over 70% and 60%, respectively. These results validate the model’s effectiveness in mining spatiotemporal correlations for regional water quality forecasting, offering a reliable tool toward sustainable water monitoring and ecosystem-based management. Full article

Most aircraft skin components are typical sheet metal components, and stretch forming serves as the primary forming process. The die face is the core foundation for both the finite element simulation and mold trial. Due to the intricate geometric characteristics of aircraft skin components and iterative revisions caused by stretch forming process adjustments and product design changes, the die face design of aircraft skin components is inherently time-intensive, highly complex, and prone to instability. To address these issues, a Hierarchical and Hybrid Association Method (HHAM) based on a robust updating mechanism and hybrid associations is proposed for the intelligent design system. HHAM can significantly enhance the stability and efficiency of die face design. Specifically, the hierarchical and automatic updating process of HHAM, incorporating robust error handling mechanisms, is the core methodology that guarantees the stability of complex and iterative die face design for aircraft skin. Moreover, the inter-module hybrid association, which integrates parametric modeling and automatic connection techniques, eliminates the instability in die face design updating caused by feature and topology variations. Additionally, robust geometric algorithms for wireframe modeling effectively improve the surface quality and generation success rate of the die face. The intelligent design system developed based on the CATIA platform has been successfully applied in two professional aircraft skin component manufacturing enterprises. Case studies and industrial application practices verify the effectiveness of the proposed system, achieving a 72.7% improvement in design efficiency and a 70.27% reduction in the risk of die face update errors. Full article

The Estimated Time of Arrival (ETA) of vessels is a vital operational indicator for voyage planning, fleet deployment, and resource allocation. However, most existing studies focus on short-distance liner services with fixed routes, while ETA prediction for long-distance tramp bulk carriers remains insufficiently accurate, often resulting in operational inefficiencies and charter party disputes. To fill this gap, this study proposes a data-driven stacking ensemble learning framework that integrates Light Gradient-Boosting Machine (LightGBM), Extreme Gradient Boosting (XGBoost), and Random Forest (RF) as base learners, combined with a Linear Regression meta-learner. This framework is specifically tailored to the unique complexities of tramp shipping, advancing beyond traditional single-model approaches by incorporating systematic feature engineering and model fusion. The study also introduces the construction of a comprehensive multi-dimensional AIS feature system, incorporating baseline, temporal, speed-related, course-related, static, and historical behavioral features, thereby enabling more nuanced and accurate ETA prediction. Using AIS trajectory data from bulk carrier voyages between Weipa (Australia) and Qingdao (China) in 2023, the framework leverages multi-feature fusion to enhance predictive performance. The results demonstrate that the stacking model achieves the highest accuracy, reducing the Mean Absolute Error (MAE) to 3.30 h—a 74.7% improvement over the historical averaging benchmark and an 11.3% reduction compared with the best individual model, XGBoost. Extensive performance evaluation and interpretability analysis confirm that the stacking ensemble provides stability and robustness. Feature importance analysis reveals that vessel speed, course stability, and remaining distance are the primary drivers of ETA prediction. Additionally, meta-learner weighting analysis shows that LightGBM offers a stable baseline, while systematic deviations in XGBoost predictions act as effective error-correction signals, highlighting the complementary strengths captured by the ensemble. The findings provide operational insights for maritime logistics and port management, offering significant benefits for port scheduling and maritime logistics management. Full article

Phishing emails continue to evade conventional detection systems due to their increasingly sophisticated, multi-faceted social engineering tactics. To address the limitations of single-modality or rule-based approaches, we propose SAHF-PD, a novel phishing detection framework that integrates multi-modal feature extraction with semantic-aware hierarchical fusion, based on large language models (LLMs). Our method leverages modality-specialized large models, each guided by domain-specific prompts and constrained to a standardized output schema, to extract structured feature representations from four complementary sources associated with each phishing email: email body text; open-source intelligence (OSINT) derived from the key embedded URL; screenshot of the landing page; and the corresponding HTML/JavaScript source code. This design mitigates the unstructured and stochastic nature of raw generative outputs, yielding consistent, interpretable, and machine-readable features. These features are then integrated through our Semantic-Aware Hierarchical Fusion (SAHF) mechanism, which organizes them into core, auxiliary, and weakly associated layers according to their semantic relevance to phishing intent. This layered architecture enables dynamic weighting and redundancy reduction based on semantic relevance, which in turn highlights the most discriminative signals across modalities and enhances model interpretability. We also introduce PhishMMF, a publicly released multimodal feature dataset for phishing detection, comprising 11,672 human-verified samples with meticulously extracted structured features from all four modalities. Experiments with eight diverse classifiers demonstrate that the SAHF-PD framework enables exceptional performance. For instance, XGBoost equipped with SAHF attains an AUC of 0.99927 and an F1-score of 0.98728, outperforming the same model using the original feature representation. Moreover, SAHF compresses the original 228-dimensional feature space into a compact 56-dimensional representation (a 75.4% reduction), reducing the average training time across all eight classifiers by 43.7% while maintaining comparable detection accuracy. Ablation studies confirm the unique contribution of each modality. Our work establishes a transparent, efficient, and high-performance foundation for next-generation anti-phishing systems. Full article

In semiconductor packaging and microelectronic manufacturing, inkjet printing technology is widely employed in critical processes such as conductive line fabrication and encapsulant dot deposition. However, dynamic printing defects, such as missing droplets and splashing can severely compromise circuit continuity and device reliability. Traditional inspection methods struggle to detect such subtle and low-contrast defects. To address this challenge, we propose MCHB-DETR, a novel lightweight defect detection framework based on RT-DETR, aimed at improving product yield in inkjet printing for semiconductor packaging. MCHB-DETR features a lightweight backbone with enhanced multi-level feature extraction capabilities and a hybrid encoder designed to improve cross-scale and multi-frequency feature fusion. Experimental results on our inkjet dataset show a 29.1% reduction in parameters and a 36.7% reduction in FLOPs, along with improvements of 3.1% in mAP@50 and 3.5% in mAP@50:95. These results demonstrate its superior detection performance while maintaining efficient inference, highlighting its strong potential for enhancing yield in semiconductor packaging. Full article