Search Results (12,235)

Aiming at the engineering problems of composite roll, pitch, and yaw pose errors of the roadheader body induced by floor undulation and geological variation, as well as insufficient anchoring accuracy caused by the incapability of traditional 1–3-degree-of-freedom (DOF) drilling–anchoring arms in dynamic error compensation during coal mine roadway excavation and bolting, this paper proposes an inverse kinematics solving method for a 5-DOF drilling–anchoring arm based on an improved Deep Deterministic Policy Gradient (DDPG) algorithm. Firstly, the modified Denavit–Hartenberg (MDH) approach is adopted to establish a full-link kinematic model incorporating body pose errors, where the drill rod length, mounting offset, and world coordinate transformation are fully considered. Secondly, an Actor–Critic dual-network architecture tailored for drilling and anchoring tasks is constructed with an 11-dimensional state space and a 5-dimensional action space. The coupling optimization between body pose errors and joint adjustments is realized by designing a hierarchical gradient reward function, a dynamic noise decay exploration strategy, and an optimal state restart mechanism. Finally, 1000 episodes of training and verification are carried out on a Python 3.10 simulation platform. The simulation results reveal that the average end-effector position error of the improved algorithm reaches 0.71 mm, and the deflection angle toward the roof is less than 1°, which outperforms the specified industrial standard. The proposed method realizes real-time compensation for dynamic body pose errors and provides crucial technical support for intelligent excavation and anchoring in underground coal mines. Full article

Rapid urbanization and increasing traffic density have increased the crash risk of vulnerable road users (VRUs), particularly pedestrians and cyclists. Meanwhile, the conventional statistical models have difficulties in capturing the nonlinear and complex nature of crash data, limiting their safety analysis effectiveness. This study utilized advanced Gradient Boosting machine learning and integrated it with SHapley Additive exPlanations (SHAP) using five years of crash data from Michigan, USA, employing a two-tiered modeling design consisting of a 4-class joint structure and binary subset frameworks. Rigorously evaluated using 10-fold stratified cross-validation to predict crash severity for VRUs, the CatBoost model had better predictive performance (AUC = 0.917) than LightGBM, Random Forest and the traditional Logistic Regression models. The analysis further indicated that prior crash actions, particularly risky crossing behaviors, are the most significant determinants of injury severity for both user groups. However, the pedestrian crash severity is strongly associated with lighting conditions and speed limits, while cyclist crash severity is more heavily influenced by intersection involvement and roadway geometry. Moreover, SHAP interaction analysis showed that the speed effect on severity significantly increases when it interacts with hazardous actions or poor visibility. The findings provide a critical insight into the implementation of effective measures and infrastructure improvements to increase the safety of VRUs. Full article

Background: Knee osteoarthritis (KOA) in people with overweight or obesity is clinically heterogeneous, reflecting mechanical loading, synovial and systemic inflammation, metabolic dysfunction, structural severity, and pain-related factors. We evaluated multidomain outcomes after semaglutide initiation and characterized clinically relevant response patterns. Methods: This retrospective, self-controlled cohort included 93 adults treated in routine rheumatology care; 88 completed a six-month follow-up. There was no untreated concurrent comparator; within-patient changes and exposure–outcome associations were therefore interpreted as non-causal. Changes in pain, Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) function, anthropometry, inflammatory markers, glycated hemoglobin (HbA1c), lipids, dose exposure, and safety were evaluated. Joint multivariable models included percentage weight loss and maximum achieved dose. Exploratory K-means phenotyping integrated dose, weight loss, body mass index (BMI) decrease, visual analog scale (VAS) and WOMAC improvement, and C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), and HbA1c reductions, with candidate solutions and resampling stability examined. Results: At six months, mean weight decreased by 10.88 kg, VAS pain score by 2.47 points, WOMAC total by 22.50 points, CRP by 2.67 mg/L, ESR by 7.62 mm/h, and HbA1c by 0.51 percentage points (all p < 0.001). Weight loss remained independently associated with WOMAC improvement, while achieved dose retained positive adjusted associations with pain, function, inflammatory, and metabolic changes. Exploratory clustering identified five clinically interpretable patterns along a broader response-intensity gradient, including high-burden multidomain response, high-dose concordant response, dose-modified intermediate response, inflammation-preserved response despite moderate weight loss, and dose-limited graded response. These observational coefficients and clusters cannot distinguish treatment-related effects from residual confounding or co-interventions. Conclusions: Substantial within-patient improvements were observed, but the uncontrolled design precludes causal attribution. The exploratory patterns may be compatible with contributions from baseline disease burden, dose exposure, mechanical unloading, and residual inflammatory variation; they do not establish weight-independent pharmacologic effects or validated patient subtypes. Prospective controlled, multicenter validation is required. Full article

Background/Objectives: This study assesses the outcomes of integrating deep learning-augmented zero echo time (ZTE DL) MRI sequences into standard MRI protocols for assessment of the hands and feet. Methods: In this single-center, retrospective study, a standard MRI protocol of hands and feet at 1.5 T was compared with the same protocol with an added ZTE DL sequence. Pathological changes in the bone, including subcortical sclerosis, osteophytes, joint space narrowing and fractures, were rated as present or absent. Indeterminate intraosseous lesions (erosions/ganglia) and indeterminate extraosseous lesions (ossicles/soft tissue calcifications) were additionally assessed on a semi-quantitative 4-point Likert scale. Diagnostic confidence was rated as low, moderate, or high. Standard MRI vs. ZTE-DL-augmented MRI comparisons were evaluated using a mixed-effects model with reader as a random effect, and X-ray vs. ZTE-DL-augmented MRI comparisons using a paired Wilcoxon signed-rank test. Interreader agreement (two readers) was assessed using Kappa statistics. Results: The cohort encompassed 59 datasets (feet = 22, hands = 37) of 40 patients with an average age of 51.79 (SD ± 11.84) years (58% women (n = 34), 42% men (n = 25)). Additional ZTE DL sequences resulted in similar findings, but with higher diagnostic confidence for assessment of bone changes when compared to conventional MR sequences (p-values < 0.05) and, overall, when compared to radiographs in a subgroup (p < 0.05 for five of six pathologic bone changes). Interreader agreement of diagnostic confidence was moderate to substantial (kappa 0.59–0.71). Conclusions: Addition of ZTE DL sequences to standard MRI protocols of hands and feet at 1.5 T demonstrated similar findings in the assessment of pathological bone changes, but with higher diagnostic confidence compared to conventional MR sequences and radiographs. However, further validation against a reference standard is required to determine diagnostic accuracy. Full article

Background: The growth hormone (GH)–insulin-like growth factor (IGF) axis is profoundly dysregulated in critical illness. GDF-15 and individual IGF-binding proteins (IGFBPs) have separately been proposed as prognostic biomarkers, but to our knowledge, no prior study has simultaneously characterized all major GH–IGF axis components and GDF-15 in the same critically ill cohort, precluding assessment of their joint intercorrelation structure. Aim: To provide the first simultaneous characterization of the intercorrelation structure among ten GH–IGF axis components and GDF-15 in a single ICU cohort, testing whether this structure is robust to adjustment for illness severity; and, secondarily, to describe admission discriminatory performance relative to APACHE II and SOFA. Methods: This was a prospective observational pilot study of 43 critically ill adults with admission (T01) measurement of ten GH–IGF axis biomarkers and longitudinal follow-up to day 15. Spearman correlations and hierarchical clustering characterized the admission intercorrelation structure; partial correlations adjusting for APACHE II and SOFA, and bootstrap confidence intervals, assessed robustness. Secondary analyses included the examination of admission discrimination (ROC/AUC), a leave-one-out cross-validated combined model, and longitudinal trajectories. All analyses were exploratory, hypothesis-generating, and unadjusted for multiple comparisons unless stated. Results: Hierarchical clustering identified a coherent cluster comprising GDF-15, IGFBP-1, IGFBP-2, and growth hormone-binding protein (GHBP), distinct from classical GH-resistance markers (GHR vs. healthy controls, GHR vs. admission) and from GH, IGF-1, acid-labile subunit (ALS), and IGFBP-3. Within this cluster, GDF-15 correlated with IGFBP-1 (ρ = 0.65, 95% bootstrap CI 0.44–0.78), IGFBP-2 (ρ = 0.50, CI 0.19–0.71), and GHBP (ρ = 0.47, CI 0.20–0.68); GDF-15 showed no correlation with classical GH-resistance markers. These correlations were essentially unchanged after adjusting for APACHE II or SOFA (partial ρ within 0.03–0.16 of unadjusted values), indicating the structure is not attributable to shared confounding by illness severity. This robustness extended to further adjustment for IL-6, age, BMI, and mechanical-ventilation duration, and results from all 45 pairwise T01 correlations were re-examined with Benjamini–Hochberg false-discovery-rate correction (9 of 11 nominally significant pairs retained q < 0.05). However, the GDF-15–GHBP correlation, unlike the GDF-15–IGFBP-1/IGFBP-2 correlations, attenuated substantially after adjustment for IL-6 and was not consistent across a brain-injury/non-brain-injury subgroup sensitivity analysis, indicating this specific link is less specific than the others. In secondary exploratory analyses, GDF-15 had the highest individual admission AUC (0.74) among biomarkers but was substantially outperformed by APACHE II (AUC 0.90) and SOFA (AUC 0.81); a combined GDF-15 + IGFBP-2 model did not improve on GDF-15 alone. Conclusions: This study identifies a severity-independent intercorrelation structure linking GDF-15 to inhibitory IGFBPs, distinct from classical GH-resistance signaling, in critically ill patients. Although the findings do not support any clinical application at this stage, further study in adequately powered, multicenter cohorts can be contemplated. Full article

In this paper, the concept of symmetry is applied to design active trajectory tracking control of autonomous distributed drive electric vehicles considering energy efficiency—that is, the construction and solution of active trajectory tracking controllers are symmetrical. This paper proposes a hierarchical control strategy consisting of upper-level control and lower-level control to improve trajectory tracking accuracy of DDEVs considering energy-efficiency characteristics. In the upper-layer control, a sliding mode active disturbance rejection (ADRC) controller is developed to control the front wheel steering angle and active yaw moment to achieve tracking of the desired trajectory, in which an extended state observer (ESO) is synthesized to estimate and compensate for internal model uncertainties and external environmental disturbances. In the lower-layer control, a multi-objective optimization algorithm based on Karush–Kuhn–Tucker (KKT) conditions is designed to realize the torque distribution control for improving energy efficiency and vehicle stability of the distributed drive electric vehicle. Finally, a joint simulation platform based on Matlab/Simulink-CarSim (version 2019) is established for simulation verification. The performances of ADRC, linear quadratic regulator controller (LQR), and model predictive controller (MPC) are compared in snake-like and double-lane-change maneuvers. Simulation results show that the proposed controller can effectively reduce motor energy consumption while maintaining trajectory tracking accuracy and handling stability. This work provides a certain engineering design solution for motion control of intelligent electric vehicles. Full article

Coordinating multiple unmanned aerial vehicles (UAVs) for cooperative missions requires agents that perceive their environment, reason about objectives, and generate joint actions. Vision–language–action (VLA) models unify these capabilities but lack a principled source of multi-agent training data and suffer from a training–inference discrepancy in closed-loop control. We propose MiniUAV-VLA, a compact centralized VLA controller for simulated multi-UAV search-and-elimination based on multi-agent reinforcement learning (MARL) expert distillation. A QMIX expert policy achieving 100% mission success generates multimodal demonstrations pairing rendered tactical map images with structured textual state prompts. A 158 M-parameter VLA model with approximately 65 M trainable parameters in the MiniMind-3V backbone and vision projection is fine-tuned with a multi-agent discrete action head that jointly predicts actions for all UAVs in a single forward pass. We identify a training–inference feature mismatch in behavior cloning and address it via prompt-end action pooling, which extracts action-relevant hidden states at the user–prompt boundary rather than after the generated response. In closed-loop evaluation with four drones and six mobile targets averaged over five evaluation seeds, MiniUAV-VLA reaches 74.4 ± 4.6% mission success against 9.4 ± 2.1% for a random policy and 16.2 ± 3.2% for an observation-limited greedy baseline. Across five independent training runs, prompt-end action pooling improves mean closed-loop success from 40.6% to 76.2% over the last-token alternative. These results support MARL expert distillation as a data-efficient route to compact multi-agent VLA control in this simulated setting. Full article

Accessibility evaluation in built environments increasingly requires performance-based methods that capture dynamic user–environment interaction rather than static compliance with dimensional standards. Existing computational approaches, including BIM-based rule checking and path-finding analysis, are effective for geometric verification but cannot quantify the physical demands imposed on users during task execution. To date, no existing framework combines task-driven motion generation with quantitative biomechanical analysis specifically for accessibility evaluation in built environments. This paper presents a modular simulation framework that integrates procedural kinematic planning with inverse dynamics analysis to enable performance-based accessibility assessment. The system generates deterministic motion trajectories using a gait-based abstraction layer and inverse kinematics, then computes net joint forces and torques via a stabilized Newton–Euler recursive algorithm. External contact forces, including ground reactions during locomotion, are modeled through a dedicated force management subsystem. The framework is implemented in Unity and evaluated in three representative scenarios: level walking, stair ascent, and stair descent. Computed knee joint forces fall within magnitude and temporal ranges reported in the biomechanical literature (peak forces of 2.5–4.0 body weights for stairs and 1.5–2.0 body weights for level walking), indicating biomechanical plausibility rather than subject-specific experimental validation. The framework offers a reproducible, extensible foundation for evidence-based accessibility design, allowing designers to detect and quantify biomechanically demanding interactions in architectural environments prior to construction, and supporting a shift from prescriptive compliance toward dynamic, human-centered evaluation. Full article

Cartilage degeneration is the hallmark pathological alteration in osteoarthritis (OA). The irreversible accumulation of fibrotic cartilage compromises joint function and accelerates disease progression. However, reliable biomarkers and therapeutic targets for cartilage fibrosis remain lacking. Through bioinformatic analysis of bulk RNA sequencing and single-cell RNA sequencing datasets, RARRES1 (retinoic acid receptor responder 1) was identified as a key biomarker associated with cartilage degeneration. The functional role of RARRES1 was investigated using a CTGF (Connective tissue growth factor) induced chondrocyte fibrosis model and siRNA-mediated gene knockdown. Subsequently, in vitro and in vivo experiments were conducted, including Western blotting, functional assays, flow cytometry, and pathological staining. RARRES1 was markedly upregulated in the damaged cartilage regions of patients with osteoarthritis, and this finding was confirmed in the chondrocyte fibrosis model. Betaine downregulates RARRES1 and promotes hyaline cartilage repair. Importantly, RGS2 was identified as a critical gene through which betaine exerts its effects on scavenging reactive oxygen species (ROS) accumulation. Our study demonstrates that betaine inhibits RARRES1, thereby appearing to upregulate RGS2 and promote ROS clearance to alleviate fibrotic changes in cartilage. RARRES1 may serve as a biomarker and a potential therapeutic target for cartilage fibrosis. Full article

In this study, we examine positive-unlabeled (PU) learning for classifying Polish texts intended for large language model training. We evaluate six PU methods, Naive, Clust, Strict-LassClust, Non-Strict-LassClust, LassoJoint, and Spy, under both classic and MVC non-SCAR labeling schemes. Across the benchmark, the best F1 reaches 0.906 on an open subtitles corpus under classic labeling and 0.900 under MVC labeling; the largest absolute gain over the Naive baseline is 0.811 F1 points on ulotki medyczne and 0.758 on plwiki in the classic setting. The experimental workflow supports multiple labeling rates, c calc, and repeated random seeds, and evaluates models with AUC, PR-AUC, and F1 as the main metrics. Our study focuses on six SpeakLeash datasets spanning different domains: plwiki, open subtitles corpus, job offers corpus, ISAP corpus, ulotki medyczne, and wolne lektury corpus. For each dataset, we construct a binary PU setup from the quality field, where HIGH is treated as positive and LOW is negative, while other values are removed. The resulting preprocessed tables retain only the selected linguistic and structural features, rescale numeric variables to the [0, 1] range, and remove the raw text column. Full article

Expansion joint anchorage zones are prone to premature cracking and fatigue deterioration under repeated wheel impact and interfacial stress concentration. Engineered cementitious composites (ECCs) are promising anchorage materials because of their tensile strain-hardening behavior, multiple fine cracking, and high deformation capacity. However, how ECC strength–ductility characteristics affect vehicle-induced stress redistribution and fatigue damage accumulation remains unclear. This study develops a material–structure–fatigue framework for ECC anchorage zones. Three PVA-ECC mixtures were tested, and their measured constitutive relationships were incorporated into a three-dimensional vehicle–expansion joint coupled finite element model validated using reported field strain data from a C50 concrete anchorage zone. Critical tensile stress histories were extracted for rainflow counting and Miner-based fatigue assessment. Results show that ECC reduced tensile stress concentration and increased tensile safety margins compared with C50 concrete. Under the defined loading scenario, the estimated fatigue life increased from 9.93 years for C50 concrete to 83.15 years for the best-performing ECC scheme. Ten-year comparative field observations supported the predicted durability trend. By linking ECC strength–ductility characteristics with vehicle-induced stress redistribution and cumulative fatigue damage, the proposed framework provides a quantitative basis for fatigue-resistant material selection and durability-oriented design of expansion joint anchorage zones. Full article

A two-hinged arch is part of a traditional structural element that can last for thousands of years. This mockup experiment tested two prestressed UHPFRC (Ultra-High-Performance Fibre-Reinforced Concrete) arches with a span of 10.32 m up to failure. Each arch was composed of two semi-arches, which were joined and slightly prestressed to ensure watertightness in the joint. The load acted on one third of the span, inducing eigenshape-like displacements. The first tensile crack appeared under the load at 60 kN, followed by shear cracks and tensile failure in an approximately opposite cross-section. The maximum load reached 170 kN with brittle failure afterwards. The behavior was successfully validated using a 3D finite element analysis and damage-plasticity material model. The mockup experiment proved that upscaling to a real ecoduct arch becomes possible, creating a unique structure with a very long service life. Full article

Efficient conservation prioritization of Huizhou historic buildings requires joint consideration of visible deterioration and value-bearing architectural elements. This study developed a deep learning-based decision-support framework using an original field dataset of 500 facade images and 2197 audited annotations from five towns in Shexian County, China. Separate YOLOv11n and YOLOv11n-seg models detected six decorative or typological element classes and segmented cracks, spalling, and stains, respectively. Model outputs were converted into a pathology severity index (PSI) based on mask-pixel-area ratios and a decorative value index (DVI) based on weighted element counts; the two indices were combined into a conservation priority index (CPI), whose weighting was examined through sensitivity analysis. On the validation set, decorative-element detection yielded a mAP@50 of 0.655, and pathology segmentation yielded a mask mAP@50 of 0.586. In a preliminary application to four held-out buildings, model-derived priority categories matched the blind ratings of three conservation experts in three cases. The framework offers interpretable evidence for preliminary screening and resource-allocation discussions, but it does not replace field diagnosis. Larger balanced datasets, external building-level validation, and metric calibration are required before regional deployment. Full article

3D reconstruction and classification of aircraft are two active research areas in optical remote sensing image processing which are of great significance for applications such as airport monitoring and intelligence analysis. The traditional approaches usually focus only on one of these two tasks, and all these methods suffer from inherent limitations. In the field of 3D reconstruction, most current methods require multiple-view images as input, which is rarely feasible in remote sensing. However, single-view 3D reconstruction is an inherently ill-posed problem. Existing methods, including voxel generation and mesh template deformation, still suffer from limited accuracy and poor shape fidelity. In the field of image classification, the existing methods are mainly based on deep learning. These methods require a large amount of labeled data, and they may also be misled by the color and texture features of the target in the dataset. In this paper, we propose a unified framework for simultaneous 3D reconstruction and classification, specifically tailored for aircraft targets in optical remote sensing imagery. The key innovations are threefold: First, we introduce the Signed Distance Field (SDF) implicit representation to build a prior-guided 3D reconstruction framework pre-trained on 3D model datasets. Second, to achieve the reconstruction process with a single image as input, we design a new joint optimization pipeline. We propose a novel dual-kernel differentiable rendering method, which is fused behind the SDF generation network for iterative optimization of the implicit code and pose parameters. Third, a gated feature fusion module is developed to combine the optimal latent vector from reconstruction with the classification backbone. This integration enables the joint output of 3D meshes and category labels within a unified loop. The resulting optimal latent code plays a dual role as a generative seed for high-fidelity 3D reconstruction and as a low-dimensional feature representation for target classification. Quantitative evaluations validate the superiority of our joint framework. Compared with the strong mesh-based competitor AtlasNet, the proposed method yields a 12.2% boost in mean F-score. In object classification, leveraging the 3D implicit geometric features boosts the performance to a peak accuracy of 97.88%, outperforming advanced remote sensing backbones such as RSMamba and EAM by 2.03% and 2.54%. Additionally, ablation studies confirm the indispensability of our key designs, revealing that our dual-task feature fusion strategy brings an absolute gain of 1.18% in classification accuracy, while omitting the clustering prior stages and the dual-kernel rendering method leads to a 30.4% and 10.1% degradation in Chamfer distance. Full article

Accurate characterization of subsurface attenuation hinges on the quality factor (Q), and high-fidelity Q models are indispensable for high-resolution imaging in exploration seismology. In principle, full-waveform inversion (FWI) offers the most accurate framework for parameter estimation, and decades of research have delivered substantial advances. Despite this, Q-FWI remains challenging in attenuating media, as velocity and Q exhibit strong crosstalk. Additionally, the sensitivity of data to Q is much weaker than to velocity, and absorption further diminishes signal strength. To address these issues, we introduce a staged FWI workflow that leverages the Adam optimization algorithm. To mitigate velocity–Q crosstalk in multiparameter inversion, we adopt a two-stage synchronous (joint–reset–joint) scheme: in stage one, velocity and Q are inverted jointly but the artifact-contaminated Q model is discarded; in stage two, velocity and Q are again inverted jointly using the updated velocity and the initial Q model. To avoid the additional trial forward simulations required by line-search step-length estimation, we use the Adam optimizer with parameter-specific adaptive updates. Numerical experiments on three synthetic models show that this workflow reduces velocity-related artifacts in the recovered Q models and improves the stability of dual-parameter viscoacoustic FWI. Full article