C₂₁ steroids are a class of steroid derivatives containing 21 carbon atoms in the parent nucleus. They have many pharmacological activities such as anti-tumor, immune enhancement, anti-oxidation, anti-epilepsy, and hypolipidemic, and demonstrating significant medicinal value. Asclepiadaceae plants are one of the main sources of natural C₂₁ steroids. According to the relevant literature, nearly 913 new C₂₁ steroidal compounds were isolated from Asclepiadaceae in the past 20 years. According to the number, properties and degree of unsaturation of the substituents on the skeleton, their structures were divided into 9 types (I-IX), of which type I was pregnane and type II-IX was deformed pregnane. Pregnane and deformed pregnane differ in spatial structure, biological activity and metabolic pathway, which makes deformed pregnane become the focus of drug development and chemical biology research. In this paper, the distribution, physicochemical properties and spectral characteristics of type II-IX modified pregnane were introduced.

Chitosan is the only amino-rich polysaccharide in nature that has characteristic of cationic polyelectrolyte, showing great attention in removing pollutants from water via adsorption. However, its stability is largely affected by solution pH, and it is necessary to accept modification to obtain stable adsorbents. This review summarized the synthesis of chitosan-based adsorbents i.e. hydrogel, composites and N-doped biochar, including their progress towards the removal of heavy metals and organic pollutants. The main factors in affecting removal ability and mechanisms were emphatically introduced, where challenges faced by chitosan-based adsorbents were also pointed out, associating with future direction. In particular, in-situ construction of N-doped biochar from chitosan not only provides a direction for high-value utilization of shell biomass, but has great potential in removing dye-based organic pollutants. It is necessary to do researches in-depth on chitosan-based N-doped biochar and composites in the field of water treatment.

Vehicle trajectory extraction on urban roads is crucial for intelligent transportation supervision, but existing techniques suffer from low detection accuracy and broken trajectories due to identity hopping. In this paper, a composite framework that fuses improved YOLOv7-tiny detection, StrongSORT tracking and Savitzky-Golay filter optimization is proposed. The framework is capable of efficiently extracting the trajectories of different vehicle targets using urban road surveillance video data collected by traffic monitoring devices. Based on experimental evaluation, the IYSSG framework performs well in three main tasks. In vehicle detection, the improved YOLOv7-tiny algorithm ensures the detection speed, while precision, recall rate and mAP@0.5 increase by 2.5%, 8.5%, and 3.7%, respectively, compared with the original YOLOv7-tiny algorithm. In terms of vehicle tracking, the StrongSORT algorithm achieves a 4.92% and 2.7% improvement in MOTA and MOTP metrics, respectively, compared with the DeepSORT algorithm. In terms of vehicle trajectory extraction and optimization, the Savitzky-Golay filtering algorithm effectively solves the problems of missing trajectory points and unsmooth trajectory due to objective factors such as video jitter and algorithmic errors, which helps the researchers to extract accurate vehicle trajectories from the traffic surveillance video for better analysis and localization of traffic problems.

In order to address the issues of low accuracy, large model parameters, and the occurrences of missed detections and false detections under complex backgrounds in the existing traditional photovoltaic panel defect detection, this paper proposes an efficient photovoltaic panel defect detection algorithm based on the RT-DETR model. Firstly, to boost detection accuracy, the FREBlock architecture is developed, which not only improves feature extraction but also enhances detection efficiency. Secondly, the CRDFP multi-scale feature fusion structure is designed to strengthen the integration of features across different scales. Lastly, the deformable attention mechanism is incorporated, which enables the model to focus on the information features of the region of interest. The experimental results indicate that the improved model achieves an average mean Average Precision (mAP) of 79.2%, an increase of 3.6 percentage points over the traditional model. Additionally, the model's parameters reduces 22.6%, and the computational load decreases 25.9%, demonstrating a high capacity for real-time detection.

A segmentation method based on self-distillation edge refinement is proposed in this paper to tackle the challenges of edge feature loss and excessive parameter redundancy encountered during semantic segmentation of remote sensing images. Firstly, a backbone network is constructed using EfficientNetB4 as the foundation. Subsequently, a lightweight edge refinement module is integrated into the self-teacher network branch. This module is designed to capture local information from intermediate feature maps while retaining the intermediate edge details filtered by shallow neural networks, with the purpose to improve the accuracy of edge pixel segmentation in remote sensing images. Finally, an adaptive multi-view vector is created to serve as a novel knowledge guide for encoder network training. This is achieves by utilizing the binary category labels of each image as the prediction matrix. The adaptive multi-view vector provides a better description of intra-class and inter-class distributions, as well as fitting inter-layer and intra-layer relationships. On the public datasets DeepGlobe and Vaihingen, the proposed method achieves an average intersection ratio of 72.4% and 83.3%, respectively. Comparative experiments demonstrate that the method introduced in this study enhances edge features while maintaining a balance among segmentation accuracy, model parameters, and inference speed. It has good feature extraction ability while lightweighting the model.

Haze interference can cause blurred structures and loss of details in remote sensing images, severely compromising the accuracy of downstream visual tasks. To address this challenge, this paper proposes a heterogeneous enhancement remote sensing image dehazing network that improves feature restoration from two perspectives: spatial structure modeling and frequency information integration. Specifically, a multi-scale asymmetric attention transformer module is designed, incorporating a direction-aware mechanism to enhance the modeling of blurred edges and texture details. In parallel, a wavelet-based adaptive high-low frequency enhancement module is constructed, utilizing Haar wavelet decomposition to separate frequency-domain information, where high-frequency and low-frequency submodules are employed to reinforce edge contours and structural representations, respectively. These two modules are embedded in the feature extraction and feature fusion stages, collaboratively addressing the limitations of traditional methods in directional modeling and high-frequency feature preservation. With low computational overhead, the proposed method achieves average PSNR/SSIM scores of 24.993 6/0.909 9 on the HAZE1K dataset and 33.180 2/0.894 2 on the RICE dataset, demonstrating significant advantages in detail restoration.

In order to solve the problem that current steel surface defect detection methods are ineffective in detecting small defects, an algorithm for detecting small defects on the steel surface that integrates Fourier convolution and difference perception is proposed. The algorithm uses CSP-FFCM to replace the BasicBlock in the backbone network, and performs convolution operations in the spatial and frequency domains to reduce the computational overhead and enhance the feature extraction capability of the network. Then, a multi-scale feature layer optimization strategy is proposed, which optimizes the allocation of computational resources while preserving fine-grained feature information to ensure that the model effectively captures the detailed information of tiny defects. Finally, a difference-aware feature enhancement module is designed to further improve the model’s detection performance of tiny defects by strengthening the feature representation capability of tiny defects. The experimental results show that the algorithm achieves mAP indexes of 83.7% and 73.1% on the NEU-DET and GC10-DET datasets, respectively, and exhibits significant performance advantages in the task of high-precision detection of tiny defects on steel surfaces.

To solve the problems of existing carbon emission prediction models such as a limited number of indicators and slow data updates, this article proposes a substation carbon emission prediction model based on the improved firefly algorithm (IFA) optimized BP neural network. Firstly, in response to the slow convergence speed and tendency to fall into local optima in the firefly algorithm (FA), teaching and learning factors are introduced to modify the firefly position update process to improve population fitness. Secondly, IFA is introduced to perform hyper-parameter optimization on the BP neural network model, and an IFA-BP neural network prediction model is constructed. Then, based on the CRITIC method, select key carbon emission indicators for the input layer of the prediction model. Finally, the prediction model is trained using the training set data to predict the carbon emissions of the substation based on the trained model. The simulation results show that compared with the three comparison schemes, the root mean square error (RMSE) of the proposed IFA-BP neural network prediction model decreases by 59.61%, 15.77% and 26.65%, respectively. The coefficient of determination (R²) increases by 5.66%, 1.46% and 1.15%. The feasibility and superiority of the substation carbon emission prediction model proposed in this paper are fully verified.

Water pollution issues in China are becoming increasingly prominent, making improvements in the accuracy of water quality prediction models crucial for effective water resource management and ecological protection. This study addresses the challenges of complex nonlinear relationship modeling and computational efficiency in multi-step water quality time series prediction by proposing an improved PatchTST model. The model incorporates three key module optimizations: 1) a lightweight CMixer encoder replacing the traditional Transformer encoder, which efficiently extracts temporal features through one-dimensional convolution and residual connections while reducing computational burden; 2) an Adaptive Mid-Frequency Energy Optimizer (AMEO) that enhances mid-frequency spectral information, improving the model’s ability to detect periodic changes in water quality parameters; and 3) a CKAHead prediction module based on Chebyshev polynomials and the Kolmogorov-Arnold representation theorem, strengthening the modeling of complex nonlinear relationships. In dissolved oxygen prediction at the Shimenzi section, the improved model achieves an MSE reduction of 12.9% compared with PatchTST and 14.0% compared with iTransformer, while maintaining a balance between computational efficiency and resource consumption. Furthermore, in generalization tests across five different monitoring sections, the model reduces MSE by approximately 10% compared with the next-best model for 48-hour forecasting tasks. Experimental results demonstrate that the improved model effectively enhances the accuracy and computational efficiency of multi-step water quality prediction, offering reference value for environmental time series analysis and water quality prediction research.

Critical node identification has been a research focus in social system, biological system, power system, and transportation system. Existing works exhibit excessive reliance on node degree, k-shell values, or their simplistic combinations while neglecting the influence of adjacent nodes and global positional information. This article proposes a multi-feature gravity model algorithm, termed as HKGM, to identify key nodes within complex networks. Specifically, the proposed scheme comprehensively considers node degree, local propagation capacity involving both first-order and second-order neighboring nodes, and introduces the global location information of nodes, aiming to construct an evaluation scheme that takes into account both the local and global properties of the network. Meanwhile, in response to the issues of algorithm complexity and computational cost in large-scale networks, this study optimizes the computational efficiency of the proposed scheme. To validate its effectiveness, simulation experiments are conducted on nine real-world datasets, comparing HKGM against nine classical algorithms. Results demonstrate that the proposed method outperforms others under evaluation metrics including the SIR propagation model, Kendall correlation coefficient, and CCDF monotonicity function. These findings confirm that HKGM achieves superior discrimination accuracy in key node identification tasks for complex networks, significantly enhancing detection accuracy.

Particle swarm optimization (PSO) algorithm, as a classical algorithm of swarm intelligence optimization, has been widely used in practice. However, in the face of different problems, it cannot make real-time adjustments according to the group status, and lacks certain flexibility, therefore, a particle swarm optimization algorithm (DVPSO) based on fusion density peak decision is proposed. For initialization, design the binary mapping of elite point set in order to improve the distribution quality of different types of particles. For velocity update, an information exchange mechanism based on density peak is constructed to balance the particle search tendency. For position updating, a dynamic two-neighborhood search strategy of step size search operator is proposed, which combines population state and optimization range changes to regulate particle movement and give consideration to search flexibility. 12 test functions were compared with PSO and 5 new swarm intelligent optimization algorithms, and 2 engineering problems were compared with 5 new intelligent algorithms. The results show that the DVPSO algorithm has better search accuracy and stability, which verifies the adaptability and good performance of the algorithm.

This paper investigates the impact of media coverage and imperfect vaccination on disease transmission by establishing an SIVR infectious disease model including an independent media information compartment M. The model analyzes the basic reproduction number, the stability of the disease-free equilibrium, the existence of the endemic equilibrium, and the conditions for backward bifurcation. The results indicate that the model exhibits complex dynamical behaviors, suggesting that media coverage and imperfect vaccination may increase the difficulty of disease control. Numerical simulations reveal bifurcation phenomena in the model, while also demonstrating that enhancing public awareness of media information and strengthening vaccination promotion can help reduce the number of infections. The study provides a theoretical foundation for understanding the role of media coverage and vaccination in disease transmission and offers references for formulating more effective public health strategies.

Aiming at the problem of insufficient accuracy of traditional Physical information neural networks (PINNs) when dealing with high-frequency features, this paper proposes a two-stage PINNs method based on adaptive weights (AWTS-PINNs) to solve partial differential equations with high-frequency solutions. This method is based on a two-stage training framework combining pre-training and fine-tuning. It introduces an activation function with the response ability of high-frequency features and integrates the adaptive mechanism of neural tangent kernels to dynamically adjust the weight of the loss function, thereby significantly improving the model’s ability to express and capture high-frequency features. The experimental results show that this method performs well in capturing high-frequency features. Compared with the existing methods such as PINNs, NTK-PINNS, RFF-PINNS and DG-PINNs, AWTS-PINNs has higher accuracy and solution efficiency. In one-dimensional and two-dimensional numerical experiments, AWTS-PINNs achieves the lowest test errors, with an accuracy reaching the order of 10^-4.

To address the initial value sensitivity and outlier interference issues in the application of traditional quantile matching estimation to multi-dimensional data, an improved framework integrating global optimization strategies and robust loss functions is proposed. By integrating principal component analysis, ordinary least squares estimation, least absolute deviation estimation, and random initialization to construct a diverse set of initial values, combined with a multi-start search mechanism and genetic algorithm for global optimization, the risk of parameter estimation getting trapped in local optima is effectively reduced. Meanwhile, the square loss function is reconstructed into an absolute loss function to enhance model robustness. Monte Carlo simulation results show that the success rate of parameter estimation of the improved global optimization method is significantly increased from 5% of the traditional method to 82%; in contaminated data with 20% outliers, the average absolute error of parameter estimation using the absolute loss function is reduced by 14% compared with the square loss function. Empirical studies on sensor signals show that the improved methods GLOBAL-MQE(L1) and GLOBAL-MQE(L2) reduce the Wasserstein distance index by 55% and 45% respectively compared with the benchmark method MQE. The proposed global optimization strategy and robust loss mechanism significantly enhance the global convergence and anti-interference ability of model parameter estimation. The improved quantile prediction sequence can effectively maintain the statistical characteristics consistent with the real signal, providing a more reliable modeling tool for engineering data distribution matching.

In this study, the ascidian-derived fungus Penicillium sp. HQ2-12 was investigated to explore the chemical diversity of its secondary metabolites and their antifungal activity against Diaporthe citri. The fungus was cultured on the solid medium of 0.9% sea salt rice for fermentation. Comprehensive techniques, including normal-phase and reverse-phase column chromatography, as well as semi-preparative high-performance liquid chromatography (HPLC), were employed for the isolation, extraction, and purification of secondary metabolites. The structures of the compounds were elucidated using nuclear magnetic resonance (NMR) spectroscopy and mass spectrometry (MS). Antifungal activity was evaluated using the microdilution broth method. A total of 16 known compounds were isolated,including five indole diterpenoids, namely rhizovarin A (1)、rhizovarin B (2)、19-hydroxypenitrem A (3)、PC-M4 (4)、10β-hydroxy-13-desoxypaxilline (5), three quinoline derivatives, namely viridicatol (6)、3-hydroxy-4-phenylquinolin-2(1H)-one (7)、aflaquinolone F (8), seven benzodiazepines, namely arctosin (9)、cyclopenin (10)、(3S)-1,4-benzodiazepine 2,5-dione (11)、(Z)-3-benzylidene-4-methyl-3,4-dihydro-1H-benzo[e][1,4]diazepine-2,5-dione (12)、7-methoxycyclopenin (13)、7-methoxycyclopeptin (14)、7-methoxydehydrocyclopeptin (15), and one sterol, namely ergosterol (16).Notably, compounds 1, 2, and 8 represent the first reported isolates from the Penicillium genus. Antifungal activity assays revealed that compounds 1-3 exhibited significant inhibitory activity against D. citri, with minimum inhibitory concentrations (MICs) of 64, 32 and 64 mg/L, respectively. Moreover, compound 6 exhibited significant inhibitory activity against multiple bacterial strains. This study was the first to report the antifungal activity of indole-diterpenoid compounds against D. citri, with compound 2 demonstrating strong antimicrobial activity and potential for further development and research.

This study focuses on investigating the functional characteristics of SiF^2-₆ as a dynamic counterion in a hydrothermally stable copper-cluster-based metal organic framework(MOF) [Cu₁₀(μ₆-O)(μ₃-OH)₁₂(dip)₄ (H₂O)₄](H₂O)₁₃H₂(SiF₆)₄(NTU-85). Single-crystal X-ray diffraction analysis reveals that the SiF^2-₆ anions in the channels are anchored via hydrogen-bonding networks formed by μ₃-OH groups of the Cu₁₀ clusters, lattice water molecules, and coordinated water molecules. Upon thermal removal of lattice and coordinated water molecules, the SiF^2-₆ anions lose hydrogen bonding interactions with these water species, adopting a disordered state while maintaining weak interactions with the Cu clusters, thereby forming a dynamic association. Within the gourd-shaped micropores (0.48~0.83 nm), these anions function as dynamic electronegative recognition sites. Under the synergistic effects of exposed OH^- groups and the dynamic SiF^2-₆ anions, this composite pore system exhibits exceptional adsorption performance at 298 K for polarizable small gas molecules, with uptake capacities of 40.0 cm³/g (C₂H₂) and 36.5 cm³/g (CO₂), and lower uptake for the gas with low polarizability or nonpolar gas, such as C₂H₄ (24.3 cm³/g), CH₄ (8.7 cm³/g), and N₂ (1.6 cm³/g) at 298 K. Remarkable selectivities of C₂H₂/C₂H₄ (12.7), CO₂/CH₄ (70.1), and CO₂/N₂ (34.7) were evaluated by the ideal adsorption solution theory (IAST), which highlights the critical role of counterion SiF^2-₆ in the capture of polarizable C₂H₂ and CO₂ via localized electrostatic interactions or dipole-quadrupole interaction. It has certain reference value for the development of polar gas separation technology.

Non-structural carbohydrates (NSC) are key carbon components for plant growth and metabolism. Plants optimize their survival strategies by regulating NSC allocation patterns between branches and leaves. In order to investigate the content characteristics of NSC in the branches and leaves of Cryptomeria japonica plantations of different ages, young (6 a), middle-aged (12 a), near-mature (23 a), mature (27 a, 32 a) and over-mature (46 a, 52 a) were selected as the research objects. The contents of NSC and its components in fresh and litter branches and leaves were determined. The results showed as follows: 1) The contents of soluble sugar, soluble sugar/starch and NSC decreased with the increase of age. The highest starch content was in 23 a. The content of NSC and its components in fresh leaves increased first and then decreased with the increase of age, and the highest content was found at 32 a. The content of soluble sugar/starch was the highest at 23 a. 2) The change trend of NSC and its component content in litter was not obvious with the increase of stand age. The contents of NSC and its components in litter increased first and then decreased with the increase of stand age, and the highest values were found at 23 a or 27 a. The content of soluble sugar/starch was the lowest in 12 a. 3) The contents of non-structural carbohydrate and its components in fresh and litter leaves were higher than those in fresh and litter branches, respectively. The content of starch in leaf litter was positively correlated with the content of nitrogen and phosphorus. There was a significant negative correlation between NSC and leaf nitrogen content in litter. 4) The ratio of the contents of NSC and its components in the fresh branches and leaves of the same age to that in the litters was greater than 1, indicating that C. japonica transfers NSC from senescent tissues to fresh tissues prior to senescence to optimize carbon resource utilization. These findings not only enhance the understanding of the carbon metabolism balance strategies employed by C. japonica stands of varying ages to sustain growth, but also provide a theoretical basis for improving the productivity of C. japonica plantations through rational regulation of stand age structure and optimization of forest stand management.

The shallow soil layer and pronounced nutrient limitations in subtropical karst ecosystems render litter decomposition and nutrient return crucial for maintaining ecological functions. Global changes involving continuous increase in nitrogen (N) and phosphorus (P) deposition may exacerbate nutrient limitations in karst regions, although the specific patterns of these impacts remain unclear. A one-year litter decomposition experiment under nutrient addition treatments was conducted to investigate the effects of N and P addition on litter decomposition and nutrient release in a karst forest. The results demonstrated that: (1) Sole N addition was found to significantly inhibit litter decomposition (-9.50%), particularly during the early decomposition phase (0-180 d), while N+P co-addition significantly enhanced decomposition (+10.46%) in the later phase (180-360 d). No significant effect was observed with P addition alone. (2) N addition was shown to significantly suppress cellulose decomposition (-21.04%) in the early phase and lignin decomposition (-19.57%) in the later phase. Conversely, N+P co-addition significantly promoted lignin decomposition (+26.46%) initially and cellulose decomposition (+20.76%) subsequently. (3) N addition inhibited the release of the majority of elements, although significant inhibition was only achieved for N, Fe, and Mn during the later decomposition stage. Combined N+P addition promoted the release of most elements, but significant promotion was only observed for K in the early stage and for C and N in the later stage. All nutrient addition treatments inhibited P release, but significant inhibition was only attained under sole P and combined N+P additions in the later stage. Sole P addition was observed to have no significant effect on the release of other nutrients. (4) Significant positive correlations were identified between decomposition rates and the release of C, N, trace elements, as well as C/P and N/P ratios, highlighting the co-limiting effects of N and P with particular emphasis on P stoichiometric regulation. The N-P co-limitation in karst ecosystems was revealed to induce suppression or null effects on litter decomposition and nutrient release under single nutrient additions, whereas synergistic enhancement was achieved through N-P co-addition. These findings emphasize the critical importance of considering N-P co-limitation in nutrient cycling studies of karst ecosystems.

Loropetalum chinense is the dominant tree species for karst vegetation restoration in Lijiang River Basin. Based on its clonal growth characteristics, L. chinense can well adapt to heterogeneous habitats. To investigate the impacts of clonal integration on leaf and root nutrient contents in L. chinense, this study employed a space-for-time substitution approach, focusing on L. chinense in dry-season habitats across different successional stages (shrub, shrub-tree, and tree stages) in karst rocky mountainous areas of the Li River Basin. Analyzed variations in carbon (C), nitrogen (N), and phosphorus (P) contents in leaves, roots, and rhizosphere soil of mother ramets, daughter ramets, and non-clonal plants. The interactions between plant nutrient status, rhizosphere soil physicochemical properties, and their responses to successional stage changes and clonal integration during the dry season were further explored. The results demonstrated that: 1) Leaf C, N, P and root C, P contents of L. chinense increased significantly with successional progression. Clonal integration significantly affected leaf water content, leaf C, N contents, and root C, P contents. Notably, non-clonal plants exhibited 5.83% and 5.24% higher root C content compared with mother and daughter ramets in the shrub stage, and 1.3% and 0.5% higher in the tree stage, respectively. Root P content of non-clonal plants in the tree stage was 120.86% and 75.64% higher than that of mother and daughter ramets. 2) Significant nutrient differences existed between leaves and roots, with water content, C, N, and P contents being significantly higher in leaves than in roots, except for occasional non-significant differences in C and P contents. The effects of succession stage and clonal integration on C and P nutrient distribution were significant. In the tree stage, the mother ramets and daughter ramets transported more C and P nutrients to the leaves than the uncloned plants, but there was no significant difference in nutrient distribution between shrub and tree-shrub stages. 3) Under combined effects of clonal integration and successional changes, plant-rhizosphere soil interaction was significantly stronger in shrub and tree stages than in the shrub-tree stage, with mother ramets showing higher interaction intensity than daughter ramets and non-clonal plants. The rhizosphere soil pH, available nitrogen (HN), total phosphorus (TP) and available phosphorus (AP) in the shrub stage were positively correlated with the nutrients of L. chinense, and the pH was negatively correlated. The rhizosphere soil organic carbon (SOC), total nitrogen (TN), HN, TP and AP were positively correlated with the nutrients of L. chinense at the tree stage. This study showed that the natural succession and clonal integration of L. chinense community had a certain influence on the leaf and root nutrients and rhizosphere soil of L. chinense. For L. chinense in shrub and tree-shrub stages, it was necessary to improve soil pH and increase the supply of N and P nutrients. The tree stage is limited by the high calcium and magnesium high pH in the karst area, and its soil pH needs to be improved to help use soil nutrients.

Using stem segments of Sedum plumbizincicola as explants and MS as the basic medium, the effects of different combinations of plant growth regulators on the disinfection, callus induction, adventitious bud induction, and rooting of S. plumbizincicola explants were studied. A rapid propagation system for S. plumbizincicola tissue culture was established. The results showed that the best disinfection effect was achieved by soaking in 75% alcohol for 30 seconds followed by immersion in 1 g/L HgCl₂ solution for 7 minutes. The combination of 0.5 mg/L NAA and 3.0 mg/L 6-BA was the most effective for callus induction. The combination of 3.0 mg/L NAA and 0.2 mg/L 6-BA yielded the best results for adventitious bud induction and proliferation, with a proliferation coefficient of 16.13. The best rooting effect for adventitious buds was 1.0 mg/L IBA, and the survival rate of transplanted seedlings exceeded 90%.