Search Results (9,191)

Mamba-based point cloud networks process serialized point tokens with state space layers and offer efficient classification, yet their performance is largely determined by the quality of local tokens produced before serialization. Existing tokenization compresses local patches via symmetric pooling, thereby discarding fine-grained geometric relationships and introducing aggregation bias under non-uniform sampling. To handle this, we propose a density-aware multi-level geometry enhancement method. A density-weighted token encoder estimates the local point density within each patch and adaptively calibrates point-wise contributions before aggregation, thus reducing the dominance of redundant dense samples. A multi-level local geometry branch extracts hierarchical coordinate-based neighborhood features directly from raw points and injects them into serialized tokens through residual fusion, compensating for geometric details lost during patch compression. Supervised contrastive learning is further adopted as an auxiliary regularizer to improve intra-class compactness and inter-class separability. Experiments on ScanObjectNN and ModelNet40 confirm the effectiveness of the approach: our method achieves 94.42 ± 0.09%, 92.11 ± 0.13%, and 87.89 ± 0.11% on the OBJ_BG, OBJ_ONLY, and PB_T50_RS variants, respectively, while ModelNet40 accuracy reaches 93.04 ± 0.12%. These results, obtained with only 12.57 M parameters, indicate a favorable accuracy–efficiency trade-off on the evaluated benchmarks. Full article

Cement-improved aeolian sand is a potentially sustainable geomaterial that has been used in engineering applications such as subgrade filling and slope stabilization in cold regions. To investigate the degradation patterns of the dynamic mechanical properties of improved aeolian sand following freeze–thaw cycles, this study conducted the Split Hopkinson Pressure Bar (SHPB) test on cement-improved aeolian sand under different freeze–thaw cycle numbers and cement content, revealing its dynamic mechanical characteristics, energy dissipation patterns, and failure mechanisms. The results indicate that, as the number of freeze–thaw cycles increased, internal microcracks progressively develop and propagate, and the dynamic mechanical properties of the modified soil gradually deteriorate. This is manifested by a gradual decrease in the dynamic peak stress, a decline in the dissipated energy density, and an upward trend in the fractal dimension of the fragment mass. Compared to the deterioration observed during the first ten freeze–thaw cycles, the magnitude of the changes in dynamic mechanical parameters decreased, indicating a gradual reduction in the overall rate of deterioration. The addition of cement enhances the frost resistance of the modified soil; however, an excessively high cement content shifted the failure mode toward brittle fracture. and the fractal dimension exhibits a decreasing trend. The conclusions of this study offer a theoretical framework for catastrophe mitigation in cement-improved aeolian sand, including freeze–thaw and dynamic loading. Full article

Accurately quantifying the spatiotemporal dynamics and driving mechanisms of anthropogenic methane (CH₄) emissions (MEs) is of great significance for achieving regional “dual-carbon” goals and global climate collaborative governance. However, existing ME inventories and macro-inversion models generally face bottlenecks such as coarse spatial resolution, lack of data update timeliness, and the inability of traditional static emission factors to capture non-linear responses. To address these issues, this study proposes an annual ME inventory enhancement framework integrating multi-source geographic and remote sensing data. This framework evaluates four advanced machine learning (ML) algorithms, including Random Forest (RF), Categorical Boosting (CB), Extreme Gradient Boosting (XGB), and Light Gradient Boosting Machine (LGBM), to construct a 0.1° high-resolution spatial grid of anthropogenic ME in China from 2019 to 2025. Furthermore, it introduces the SHapley Additive exPlanations (SHAP) framework and multi-scale spatial autocorrelation analysis to parse the driving mechanisms and clustering patterns. The results show the following: (1) LGBM exhibits the optimal comprehensive estimation accuracy (R² = 0.938, RMSE = 3.707 Kt) and robust capability in capturing extreme ME sources (${\mathrm{R}}_{\mathrm{Top}}^2$ = 0.929). (2) SHAP attribution reveals that coal mining and nighttime light (NTL) represent the primary contributing features to ME predictions (with a cumulative contribution of 65.60%), followed by agricultural and pastoral activities (24.98%), and all factors exhibit significant non-linear threshold and step-response characteristics. (3) Regarding spatiotemporal evolution, China’s total anthropogenic ME shows a trend of initial slow increase followed by high-level stabilization; spatially, it presents a “hot in the north, cold in the south” pattern, with extreme high values highly clustered in the Shanxi–Shaanxi–Inner Mongolia energy triangle and its peripheral expansion nodes. This study provides scientific references for formulating tailored, multi-scale, and refined CH₄ mitigation strategies. Full article

To reduce energy consumption in mine ventilation networks under dynamic air demand, this study proposes an Improved Cuckoo Search (ICS) algorithm for energy-efficient airflow regulation. A nonlinear optimization model is established incorporating airflow balance at nodes, air pressure balance in loops, and fan operating constraints; equality constraints are handled by an exterior penalty function and inequality constraints by an interior penalty function. To overcome standard Cuckoo Search’s insufficient convergence accuracy and premature convergence, adaptive step sizes, an adaptive discovery probability adjustment, and a hybrid update strategy with differential evolution are introduced to enhance global exploration and local exploitation, convergence speed, and stability. The method was validated on an intelligent ventilation platform using field data from the Second Panel of a coal mine in Shaanxi Province. Results show that ICS reduces total power consumption from 482.0 kW to 424.9 kW, achieving an energy-saving rate of 11.85%, and outperforms CS, PSO, and GWO in convergence efficiency, accuracy, and stability. Quantitatively, it converges in 130 iterations (190 for CS, 160 for GWO), with a 0.038 kW objective-function standard deviation (coefficient of variation 0.009%) over 30 runs and 0.22 s per run. The proposed method optimizes airflow distribution while satisfying air-demand constraints, offering a practical approach for energy-efficient mine ventilation. Full article

Promoting visual health in early childhood is crucial for visual development, and effective school-based interventions are urgently needed. This study examined the effects of physical exercise combined with accommodative training on visual acuity and ocular biometric parameters in children aged 6–7 years. A total of 163 first-grade students were allocated by intact classes to an experimental group (n = 82), which received a 32-week physical exercise program incorporating visual tasks designed to stimulate ciliary-muscle accommodation, or a control group (n = 81), which received standard physical education. Uncorrected distance visual acuity (UDVA), kinetic visual acuity (KVA), axial length (AL), corneal curvature (CC), and anterior chamber depth (ACD) were assessed at baseline, mid-intervention, and post-intervention. Significant time effects and time × group interactions were observed for UDVA and KVA (all p < 0.05). Although no between-group differences were found at baseline or mid-intervention, the experimental group demonstrated significantly better UDVA and KVA than the control group after the intervention. Significant time effects were observed for AL, CC, and ACD, but no significant between-group differences were detected. These findings suggest that physical exercise combined with accommodative training can effectively enhance visual function in young children, while ocular biometric parameters maintain stable growth trajectories consistent with normal pediatric development. Full article

Litter size is an economically important reproductive trait in dairy goats, but its genetic improvement remains challenging because of its complex genetic architecture and the influence of multiple environmental factors. This study aimed to identify genomic loci associated with litter size-related traits and to evaluate the potential of an intergenic variant near the candidate gene CTNND2 as a molecular marker in Saanen dairy goats. A total of 341 does with reproductive records were analyzed using a genome-wide association study (GWAS), followed by candidate SNP genotyping and association analysis. A shared association signal for litter size and live-born kid number was identified on chromosome 20, leading to the selection of the intergenic SNP CM048886.1:g.61357858G>A located near CTNND2 for further evaluation. Three genotypes (GG, AG, and AA) were detected, and the locus exhibited moderate genetic diversity (PIC = 0.27) while conforming to Hardy–Weinberg equilibrium. Association analysis revealed nominal associations between the SNP and both litter size (p = 0.0445) and live-born kid number (p = 0.0448), whereas no significant association was observed with average kid birth weight (p = 0.125). Goats carrying the A allele showed numerically higher litter size and live-born kid number than those carrying the G allele. These findings identify CM048886.1:g.61357858G>A, an intergenic variant near CTNND2, as a promising candidate marker for litter size-related traits in Saanen dairy goats and provide a foundation for future validation studies and the development of molecular breeding strategies to improve reproductive performance. Full article

The increasing frequency of wildfires has made large-scale collective emergency evacuation increasingly critical. However, existing studies provide limited understanding of how information structures shape the interaction between individual risk perception and collective evacuation behavior. This study develops a collective evolutionary game-based evacuation framework under ambiguous and clear information conditions. Under ambiguous information, individual heterogeneity in risk sensitivity, mobility, and resource endowment is incorporated into social interaction payoffs. Under clear information, observable evacuation consequences, including travel time, risk exposure, and congestion effects derived from route-choice interactions, are incorporated into evacuation utility. Numerical simulations examine the evolutionary characteristics of collective evacuation behavior under different information conditions and population scales. The results show that social interactions play an important role in shaping evacuation decisions under ambiguous information, while congestion effects and route-choice interactions influence evacuation utility under large-scale demand. Sensitivity analyses further demonstrate that congestion representation affects evacuation utility across different population scales. These findings highlight the importance of considering information structure, individual heterogeneity, and collective interactions in evacuation modeling. Emergency management should therefore improve risk communication, evacuation capacity, and congestion mitigation strategies. This study provides theoretical insights into collective evacuation decision-making under heterogeneous information conditions. Full article

Toxicodendron vernicifluum is an economically important lacquer-producing tree endemic to East Asia; propagation bottlenecks severely restrict elite clone industrialization. This review summarizes 20 years of the domestic and international literature on seed propagation, root cutting, stem cutting, grafting, and tissue culture of lacquer tree, compares technical performance among China, Japan, Vietnam, and Mediterranean Rhus species, and quantitatively summarizes germination/rooting rates across protocols. Sexual propagation has low rooting efficiency; tissue culture faces high cost and browning obstacles; molecular regulatory mechanisms remain poorly characterized. We propose an integrated breeding framework combining molecular mechanism dissection, low-cost vegetative propagation optimization, and standardized plantlet production. Future directions including multi-omics analysis, gene-edited easy-rooting germplasm, and bioreactor micropropagation are highlighted to support large-scale elite clonal stock production in Qinba Mountain regions. Full article

Object overlap in X-ray security inspection images can introduce interference into feature representation and increase the difficulty of prohibited item detection. To address this problem, an improved RT-DETR model named RPR-DETR is developed in this study. First, the original backbone is replaced with a lightweight RGCSPELAN structure, where structural reparameterization is incorporated into hierarchical feature aggregation to decrease model complexity. Second, a Pola-CGLU encoder is introduced to model relationships across different spatial regions while retaining local neighborhood interaction. In addition, an RFPN structure is employed to reorganize feature transformation and scale conversion among different feature levels. On the OPIXray dataset, RPR-DETR obtains an mAP50 of 90.0%, exceeding RT-DETR-R18 by 1.2 percentage points. Compared with RT-DETR-R18, RPR-DETR uses 31.2% fewer parameters and requires 22.6% fewer FLOPs. On DvXray, the proposed model improves mAP50 by 1.4 percentage points over the baseline. The experimental results indicate that RPR-DETR provides a better balance between prohibited item detection performance and model complexity. Full article

In the past decade, the green transformation of agricultural land use has been essential for addressing agricultural non-point source pollution while safeguarding food security. However, existing policy interventions often suffer from a persistent “high investment–low responsiveness” paradox, reflecting a structural mismatch between top-down policy supply and farmers’ demand. Using survey data from 478 rural households in Shaanxi Province, China, this study applies a choice experiment and mixed logit model to examine farmers’ preferences for green land-use incentive policies, preference heterogeneity, as well as policy interaction effects across production-end and sales-end dimensions. Three main findings emerge. First, farmers exhibit a hierarchical preference structure characterized by sales-end prioritization and production-end support as a safety net. Green government procurement and green product certification generate significantly higher willingness to accept (WTA) than conventional production subsidies, while green credit support is the least preferred instrument due to strong risk aversion. Second, preferences vary systematically across educational attainments and geographic regions: less-educated farmers favor direct technical guidance and input support, whereas more-educated farmers are more receptive to green credit; geographically, farmers in Shangluo exhibit pronounced caution toward debt-based credit, while those in Hanzhong and Ankang display stronger demand for green inputs and technical services. Third, policy combination effects are highly differentiated: within the production domain, subsidies complement technical guidance; within the sales domain, procurement and certification exhibit a substitution effect; across domains, significant complementarities emerge when certification is paired with production subsidies or green input provision, whereas credit support and certification act as substitutes. These findings reveal the heterogeneous structure of farmers’ policy preferences and the interactive mechanisms among policy instruments. The study provides a demand-oriented empirical basis for designing targeted and coordinated policy portfolios to promote green land-use transformation for similar ecologically sensitive mountainous agricultural regions in China. Full article

Long-term distribution planning requires weather sequences that capture not only changes in temperature and precipitation, but also the dependence and persistence among heat, humidity, wind, solar radiation, and rainfall. This paper develops multivariate weather scenario generation for Shaanxi Province, China, using daily data from six NEX-GDDP-CMIP6 models for 1985–2014 and 2031–2060 under SSP2-4.5 and SSP5-8.5 at four locations. Two generators are compared: a vector autoregressive model with seven-day residual blocks, and a season-conditioned multivariate nearest-neighbor analog. The comparison leaves out complete five-year periods within 24 climate-model–location units and tests whether the future-minus-historical changes in 16 weather and compound-event indices are preserved. The two methods each obtain the lower overall loss in 12 units. The vector autoregressive method better preserves most marginal and correlation signals, whereas the nearest-neighbor method better preserves RX3day, hot–dry–low-wind days, dry-spell duration, and three-day compound stress. Both are retained to generate 960 30-year paths, which drive a fixed IEEE 33-node resilience example showing that generator differences propagate to demand, renewable availability, repair duration, and unserved energy. The resulting conditional scenario sets provide an auditable weather basis for climate-resilient and sustainable long-term planning of distribution systems, which is a prerequisite for a sustainable energy transition under a changing climate. Full article

Commercial sesame paste exhibits substantial variations in color and aroma, reflecting differences in raw materials and commercial processing conditions, including roasting. This study investigated 15 commercial sesame paste products to characterize associations among color/browning characteristics, sensory attributes, and aroma-active chemical profiles. Across the commercial samples, darker color and higher browning index ($BI$) values were generally associated with stronger roasted nutty and burnt attributes and weaker grassy notes. Higher pyrazine levels were also observed in samples exhibiting more pronounced apparent roasting-related aroma characteristics and were positively associated with roasted nutty and burnt attributes. GC-O and OAV screening, together with exploratory PLSR further analysis, indicated that Strecker aldehydes, pyrazines, and phenolic compounds were candidate contributors to the observed sensory differences. In aroma addition experiments, under the tested addition conditions, the R mixture containing 2-methylbutanal and selected pyrazines enhanced roasted nutty intensity, whereas the B mixture containing guaiacol and 4-vinyl-2-methoxyphenol enhanced burnt/smoky intensity under the tested addition conditions. These findings demonstrate that color/browning characteristics and aroma-active compound profiles provide complementary information for characterizing apparent roasting-related differences and aroma quality in commercial sesame paste, without representing direct measures of controlled roasting degree. Full article

Under China’s dual-carbon strategic goals, large-scale public centralized heating plays a critical role in energy conservation. However, traditional manual open-loop control suffers from high response latency. Furthermore, existing unidirectional predictive methods lack dynamic feedback correction mechanisms. To address these issues, this study proposes an intelligent dual-system collaborative control architecture specifically designed for public heating systems. This architecture utilizes a cloud–edge device framework. It establishes a two-way linkage between heating equipment and the cloud decision platform. Consequently, it constructs an integrated regulation framework encompassing forward decision generation, reverse state verification, and dynamic feedback correction. Specifically, the forward module utilizes a Mamba-structured state-space model to generate data-driven boiler operation strategies. Meanwhile, the reverse module employs a Temporal Convolutional Network with Monte Carlo Dropout (TCN-MC Dropout). This probabilistic network enables state inversion evaluation with reliable uncertainty prediction intervals. These two modules are deeply coupled through an adaptive feedback correction mechanism. Together, they significantly improve system stability and operational robustness under complex thermal disturbances. Specifically, the proposed architecture achieves a room temperature compliance rate exceeding 96% and restricts temperature fluctuations to within $\pm 0.75 °\mathrm{C}$. Simultaneously, it reduces boiler energy consumption by over 16.4%. This solution has been successfully deployed in the heating network at Shaanxi Normal University as a representative real-world case study. Ultimately, it provides a practical technical reference for the intelligent upgrading and low-carbon transformation of public centralized heating systems. Full article

The adult heart relies on fatty acid β-oxidation for energy, and cardiomyocyte lipid droplets (CLDs) are central to myocardial lipid homeostasis. CLDs are now recognized to protect cardiomyocytes against lipotoxicity, endoplasmic reticulum (ER) stress, oxidative stress, and nutrient deprivation. Moreover, accumulating evidence suggests that CLDs within the same cardiomyocyte are not uniform but may exhibit heterogeneity in size, subcellular localization, protein composition, and metabolism. Here, we propose a conceptual framework that classifies CLD heterogeneity along three dimensions: morphology, organelle connectivity, and molecular composition. We first define the biological characteristics of CLDs and categorize their heterogeneity. We then discuss the molecular pathways maintaining this heterogeneity, their roles in the healthy heart, and their dysregulation in cardiovascular diseases. Finally, we review preclinical strategies targeting CLD heterogeneity and outline challenges and future directions for clinical translation. This framework integrates current evidence while acknowledging that direct molecular validation awaits future single-droplet omics technologies. Full article