In order to solve the problem of chaotic time series prediction of permanent magnet synchronous motors, a model containing two linearly coupled reservoir computers is proposed in this paper. By seeking the optimal coupling coefficients in the two reservoir computers, the chaotic behavior of the target system can be predicted based on the time series data of the target system, while achieving higher prediction accuracy and longer prediction time. Using a PMSM as the object of study, it is demonstrated that the linearly coupled reservoir computers models give the best prediction with an RMSE=9.14×10^-4 at a coupling factor α=0.803.

To solve the problem of simple behavior and small key space in low dimensional chaotic systems, a design method of multi-dimensional chaotic systems is proposed. The general formula of system structure and equilibrium point are given, and the chaotic property of Smale horseshoe system is verified by Si’lnikov theorem. Taking a three-dimensional chaotic system as an example, its dynamic characteristics are analyzed and its displacement control is enhanced. The chaos system is constructed by Multisim circuit simulation software, and the realization of the system is verified. Based on the adaptive theory, a multi-channel adaptive synchronization controller is designed. The controller can realize synchronization under various system error combinations and improve the security of chaotic systems in secure communication.

The existing Fritillaria ussuriensis Maxim topsoil stripping machine is not highly intelligent and may cause Fritillaria ussuriensis Maxim damage when harvesting Fritillaria ussuriensis Maxim, so it is necessary to develop an intelligent Fritillaria ussuriensis Maxim topsoil stripping machine, and the first step to realize intelligent topsoil stripping is to realize intelligent identification. Therefore, a Fritillaria ussuriensis Maxim detection model based on improved YOLOv5-YOLOv5-Swin-L is proposed. Swin Transformer is introduced to replace C3 in the backbone network to reduce the sequence length and computational complexity, thus simplifying the number of parameters in the model. At the same time, the activation function in the original network structure is replaced by ACON activation function, which can improve the accuracy and robustness of the model. Experimental results show that the improved YOLOv5-Swin-L can identify Fritillaria ussuriensis maxim with the highest accuracy of 96.39% and the highest recall rate of 95.76%, which is superior to other network models of YOLOv5 series.

To address the issue of insufficient accuracy when solving convection-dominated diffusion equations with drastic gradient changes near the boundary layer using Physics-Informed Neural Networks (PINNs), a neural network solution method incorporating a parameter progressive strategy is proposed. This method initially approximates the smooth solution of the large diffusion parameter equation, then progressively reduces the diffusion parameter while using the optimal network parameters from the large diffusion parameter as the initial values for training the small diffusion parameter neural network. By iteratively optimizing the Physics-Informed Neural Networks through parameter cycling, the neural network’s representational capacity is enhanced, thereby improving the approximation accuracy of the convection-dominated diffusion problem. And finally a high-precision singular solution for the small diffusion parameter is obtained. A comparison of the accuracy and convergence efficiency of present method with those of PINNs and gPINNs shows that this method can efficiently approximate the large gradient solution of the convective dominant diffusion equation with an accuracy of the order of 10^-3 under the condition of unknown boundary layer position. Meanwhile, the present method has more advantages and better performance than PINNs and gPINNs in terms of convergence speed and stability.

Web application attacks have always been a hot issue in cyberspace. With the continuous development of Web attack techniques, traditional intrusion detection systems and Web application firewalls are increasingly unable to meet the security protection needs. A Convolutional Gated Recurrent Unit (CGRU) neural network for detecting malicious Web requests based on feature fusion is proposed in this paper. The local features and high-order features of network events were captured by the designed network using CNN, traditional pooling methods are abandoned, and GRU is employed to collect features in the time dimension, replacing the original pooling layer. In addition, to improve detection performance, nine traditional machine learning statistical features are selected to complement the original features that perform effectively in web attack detection. Furthermore, the Word2Vec model is utilized to pretrain the word embedding matrix and obtain the input of CGRU model, which enables final results to be classified and facilitated the improvement of multi-classification accuracy. The proposed method is compared with current typical methods on the public HTTP CSIC 2010 dataset, and the results show that the accuracy, recall, F1-score, and precision of the proposed method are 99.81%, 99.78%, 98.80%, and 99.81%, respectively, which are all improved compared with the existing methods.

The purpose of this paper is to investigate argument extraction techniques, in order to identify, extract, and analyze argumentative components and structures in textual information. The intelligent analysis of debate fact text is accomplished by extracting arguments related to the topic of debate from multiple sentences and determining whether the position of the argument is supportive or oppositional. Previous research has mainly relied on deep learning models such as convolutional neural networks and recurrent neural networks, which have simple network structures and cannot learn deeper features from arguments. In order to learn richer semantic information from argumentative text for position classification better, this paper proposes an enhanced RoBERTa model (EnhRoBERTa) based on the pre-training language model RoBERTa, which fully utilizes the multi-level multi-head attention mechanism and extracts shallow and deep semantic representations for fusion, enabling a comprehensive understanding of the relationship between arguments and debate topics from multiple feature dimensions, thereby facilitating argument position classification. However, considering the problem of imbalanced distribution of position in argumentative points, this paper adopts data augmentation techniques to enhance the learning ability of scarce samples. The experimental results on the CCAC2022 match data set show that the proposed model can extract more text features than other baseline models, achieving an F1-score of 61.4%, which is approximately 19% higher than that of the baseline models TextCNN and BiLSTM, and 3.8% higher than that of the RoBERTa.

Chinese idioms have unique abstract meanings. In machine reading comprehension, in order to solve the problem that the model cannot fully understand the complex semantics of Chinese idioms, a cloze-test matching network is proposed. The matching network incorporates the idioms and their dictionary interpretations into the model in an attentional interaction manner, so that the idioms get a better idiom vector representation, and a bidirectional interaction strategy is adopted between the passages and the candidate answers. A model is designed for the cloze-test reading comprehension task using matching networks combined with BERT and ERNIE language models, respectively. The model outperforms other traditional models, SKER model and BERT model combined with enhanced global attention, and achieves 77.0% accuracy on the Chinese idiom dataset CHID.

Knowledge distillation can improve the generalization ability of neural networks and solve the problem of insufficient labeled data when classifying remote sensing image scenes. And the high similarity between classes existing in remote sensing images can lead to the loss of intermediate knowledge features. To address this problem, a feature extraction method (SDCASA) based on the self-distillation cascaded attention mechanism is proposed. Firstly, a teacher and student network with shared weights is constructed; then the cascaded attention module is used to refine the features extracted by the deep teacher network while retaining the intermediate edge information filtered by the shallow neural network. Secondly, the refined features are used to guide the student network to learn. Finally, a linear classifier is trained downstream to complete feature classification. The classification accuracies of 85.17%, 90.10%, 91.13% and 85.50%, 92.13%, 91.17% are achieved on three publicly available datasets AID, MLRSNet, and EuroSAT using 20% and 50% of the samples trained, respectively. This method can effectively improve the classification accuracy of remote sensing image scenes and outperforms the mainstream self-supervised image classification methods SimSiam, SwAV, MoCov2, Deepcluster, and has good application value.

A class of nonlinear p-Laplacian fractional differential equation coupling systems with multipoint boundary value problems is considered where the nonlinear function contains the Caputo fractional derivative and the boundary conditions include nonlinear integral terms. Based on the generalized fixed point theorem of sum operator and the properties of fractional calculus operator, the unique positive solution of the coupling system is analyzed. The existence of the unique positive solution is deduced by means of the corresponding operator equation, and the main results are obtained. The main results are tested by numerical examples.

In this paper, the existence of solutions for a class of Kirchhoff type problems with critical exponent and indefinite potential is studied. Firstly, the energy functional is proved to satisfy the mountain pass geometry. Secondly, the compactness condition of function is obtained by proving that the energy functional satisfies the local (PS)_c condition. Finally, the existence of two nontrivial solutions is obtained via Ekeland’s variational principle and the mountain pass lemma.

In this paper, the following concave-convex fractional Schrödinger-Poisson system with doubly critical exponents is investigated
$\left\{\begin{array}{lr}(-\Delta)^s \boldsymbol{u}-\boldsymbol{\phi}|\boldsymbol{u}|^{2_s^*-3} \boldsymbol{u}=|\boldsymbol{u}|^{2_s^*-2} \boldsymbol{u}+\lambda h(\boldsymbol{x})|\boldsymbol{u}|^{q-2} \boldsymbol{u}, & \boldsymbol{x} \in \mathbf{R}^3, \\ (-\Delta)^s \boldsymbol{\phi}=|\boldsymbol{u}|^{2_s^*-1}, & \boldsymbol{x} \in \mathbf{R}^3,\end{array}\right.$
where 1<q<2, s∈(0,1), λ>0 is a real parameter and h satisfies some certain conditions. It is showed that there exists λ^*>0 such that the system has a positive local minima solution with negative energy and a positive mountain-pass solution with positive energy for any λ∈(0,λ^{*) in Ds,2(R3) by applying the Mountain Pass Theorem and variational method.}

The article investigates the isomerisation reaction of clusters based on the topological principle, using density general function theory, from the perspective of thermodynamics and kinetics, and finally obtains the following conclusions: the process of isomerisation reaction of clusters CrPS₄ mainly depends on the breaking and formation of bonds between non-metallic atoms. The better the stability of the products of the isomerisation reaction, the greater the difference in stability between the reactants and the products, the greater the equilibrium constant of the isomerisation reaction and the more completely the reaction proceeds. The more unstable the reactants are, the higher their energy content and the lower the activation energy required to leap to the activated state, the more likely the isomerisation reaction will occur. All configurations are converted to the more stable configurations 1⁽⁴⁾, 1⁽²⁾, 3⁽⁴⁾, 2⁽²⁾ and 2⁽⁴⁾ by isomerisation reactions. 5⁽⁴⁾-2⁽⁴⁾ and 6⁽⁴⁾-1⁽⁴⁾ are the fastest and most complete reactions. The entropy-increasing positive reaction has a greater finger front factor than the reverse reaction finger front factor, contributing more to the positive reaction rate, and this difference makes the isomerisation reaction easier to carry out.

To construct the porcine PPARD gene editing vector and detect the gene editing efficiency. In this study, two sgRNA sequences that can target the cutting of PPARD gene were designed according to the structure and sequence characteristics of PPARD gene, and the two PPARD sgRNA expression boxes were connected to a CRISPR/Cas9 vector in tandem. After the constructed plasmids were transfected into PK15 cells, the DNA of each group of cells was extracted. The mutant region was amplified by PCR, and the vector editing efficiency was detected at the DNA level by sequence determination.The total RNA and total protein of each group were extracted, and the gene editing efficiency of recombinant vectors was detected by qRT-PCR and Western blot at the level of gene expression and protein expression. The total mutation rate of the PPARD-2KO group was 45.83% (22/48). Compared with the normal control group, the PPARD mRNA in the PPARD gene editing group was significantly decreased by 84% (P<0.01), and the PPARD protein expression was significantly decreased by 61% (P<0.01). The results show that this study successfully constructed a high-efficiency PPARD gene editing vector by using the improved CRISPR/Cas9 vector system, and can achieve high-efficiency gene editing on cells without drug screening, which will greatly improve the efficiency of subsequent screening of single-cell cloning and promote the functional study of PPARD genes.

Various pathogens that live or are parasitic in the wild may have the potential to infect humans through evolution, genetic variation, or ecological suitability alteration. Therefore, it is important to carry out risk analysis of spatial and temporal intersection of human and important host animals for traceability tracking of unknown epidemic infections, field environmental health management and early warning. Bats carry a variety of dangerous pathogens. In this paper, a random forest model was constructed to predict the suitable distribution area of bats as the target hosts, and spatial variables related to bats and human activities were integrated to establish a risk assessment model with spatial intersection of humans and bats, and to produce a risk distribution map; for the convenience of administrative management, a risk analysis at the township level as the management unit was conducted. The results show that, 1) The dominant variables affecting bat fitness are average annual precipitation, annual temperature fluctuation range, monthly mean diurnal temperature difference, seasonal variation of temperature, seasonal variation of rainfall and mean temperature of the coldest quarter. 2) The random forest predicted the high fitness distribution area of bats was 9.796 5×10⁴ km², accounting for 24.86% of the total area of Yunnan Province, mainly distributed in southern part of Yunnan Province. 3) The risk assessment of epidemic infection found 45 townships belonging to the high risk area of epidemic infection, with an area of about 0.534 6×10⁴ km² and an affected population of about 2.87 million, mainly distributed in the southern, southwestern and central-eastern parts of Yunnan Province.

In order to explore the hypoglycemic effect of endophytic fungi from the mangrove Bruguiera gymnorrhiza, the endophytic fungus Penicillium sp. GXIMD00006 was fermented and isolated by various column chromatography techniques. By using MS, NMR and literature comparison,11 compounds were identified as (+)-sclerotiorin (1), sclerotioramine (2), isochromophilone VIII (3), terretonin (4), terretonin A (5), chrodrimanin A (6), chrodrimanin B (7), chrodrimanin E (8), (22E,24R)-3β,5α-dihydroxy-ergosta-7, 22-dien-6-one (9), erogosterol (10) and ergosterol peroxide (11). Besides, terretonin (4) and terretonin A (5) were first isolated from the Penicillium sp. The α-glucosidase inhibitory effects showed that (+)-sclerotiorin (1) had significant inhibitory activity with IC₅₀ of 3.30 μmol/L, which indicated that the secondary metabolites of endophytic fungi from the Bruguiera gymnorrhiza had the potential to be developed into new α-glucosidase inhibitors.

The presence of antibiotic pollutants in wastewater treatment process may affect the microbial activity in sludge, and the microbial activity directly affects the biological phosphorus removal effect of wastewater. Therefore, the effect of ciprofloxacin (CIP) on the operation efficiency of the anaerobic reactor treating high-concentration phosphorus wastewater and the response of microbial community were explored, and the COD removal rate, total phosphorus and phosphate removal rate, related enzyme activity and microbial community structure wereinvestigated. The results showed that the COD removal rate of high-concentration phosphorus wastewater in the reactor was stable at about 98% without CIP, and the average COD removal rate decreased to 63.1% when the CIP concentration was 10 mg/L, and the total phosphorus removal rate showed a downward trend in fluctuations. The addition of 10 mg/L CIP had a great effect on the concentration of phosphine; peaks associated with tryptophan proteins and soluble microbial byproducts and coenzyme F₄₂₀ appeared in the three-dimensional fluorescence spectra of LB-EPS, and humic acid-like peaks appeared in TB-EPS. Chloroflexi was the dominant phylum under CIP stress, and its relative abundance were 43.13%, 44.48% and 39.96% when the CIP concentration in the influent were 3 mg/L, 5 mg/L and 10 mg/L, respectively. Methanobacterium and Methanothrix were the dominant genera of archaea, thus ensuring the efficient operation of the UASB reactor. KEGG functional analysis showed that archaea and bacteria were mainly carbohydrate metabolism and energy metabolism. After increasing CIP concentration, the functional gene abundance of membrane transport decreased greatly, from 10.15% to 9.31%, indicating that increasing CIP concentration had a greater impact on archaea.

Fixed-point sampling was conducted in the field at different periods to explore the change of carbon, nitrogen, phosphorus, potassium and the stoichiometry characteristics in the roots, stems and leaves in different phenological periods of the Alchornea trewioides which is known as the pioneer plant in karst area. The contents of soil carbon, nitrogen, phosphorus and potassium were measured and the content of total carbon (TC), total nitrogen (TN), total phosphorus (TP) and total potassium (TK) in plants and their relationships with soil nutrients were analyzed, and the correlation between plant element content and basic soil nutrients content were analyzed. The results showed that: (1) From florescence to abscission period, the variation range of TC, TN, TP and TK are 428.42~460.64 g/kg, 8.43~25.81 g/kg, 0.37~1.13 g/kg, and 2.82~6.32 g/kg, respectively The contents of TC, TN, TP and TK in leaves are the highest in all organs. At florescence, the contents of C, N, P and K in leaves were 446.75, 24.12, 1.13 and 6.32 g/kg, respectively. At fruit stage, the contents of C, N, P and K in leaves were 451.405、25.805、0.88 and 5.54 g/kg, respectively. In the abscission period, the contents of C, N, P and K in leaves were 460.635、19.02、1.07 and 5.52 g/kg, respectively. The contents of C, N and K in roots are more stable than those in stems and leaves, and the fluctuation of P content in roots is larger than that in stems and leaves. The variation coefficient of TC, TN, TK in different organs is stem >leaf > root, and the variation coefficient of TP is root > leaf > stem. The variation coefficient of TP in root is 39.1%, 250.04% higher than that of stem and 206.67% higher than that of leaf. (2) From florescence to abscission period, the variation range of m_C∶m_N, m_C∶m_P, m_C∶m_K and m_N∶m_P in roots, stems and leaves were 17.53~51.99, 396.86~1 179.7,71.16~154.88 and 13.28~29.59, respectively. Among all organs, the mass ratio of m_C∶m_N, m_C∶m_P and m_C∶m_K of leaves is the lowest. At florescence, the mass ratio of m_N:m_P is root > leaf > stem, in fruit and abscission period it is leaf > root and stem. The mass ratio of m_N∶m_P of roots, stems and leaves approaches 16 in abscission period.The m_N∶m_P mass ratio of the leaves of Alchornea trewioides were higher than 16, and the growth of the Alchornea trewioides is limited by P. The interaction effects of different organs and phenophases on m_C∶m_N, m_C∶m_P, m_C∶m_K and m_N∶m_P were significant. (3) Soil total nitrogen had an important effect on plant TN content, and soil available potassium had an important effect on plant TC and TK content, while soil total carbon and available phosphorus content had no significant effect on plant TC, TN, TP and TK.

The changes in soil physicochemical properties caused by changes in litter input can significantly affect the content of soil carbon, nitrogen, and phosphorus nutrients, as well as changes in soil extracellular enzyme activity and stoichiometry characteristics. To further clarify the impact of changes in litter input on soil nutrient content, extracellular enzyme activity, and stoichiometry in karst areas, and to explore the ecological response patterns between extracellular enzymes and environmental factors, in this study, the soil of Loropetalum chinense aged forest in karst area was taken as the research object, and litter addition and removal experiments were arranged. Through the determination of soil nutrients and extracellular enzyme activities, the theory and methods of Ecological stoichiometry were used to systematically study six extracellular hydrolases in the soil(β-1,4-glucosidase (BG), β-1,4-Xylosidase (BX), cellulose hydrolase (CBH), β-1,4-acetylglucosaminidase (NAG), leucine aminopeptidase (LAP), acid phosphatase (AP) and two extracellular enzymes (catalase (CAT) and polyphenol oxidase (PPO)), as well as their stoichiometric effects, were studied, and their correlation with soil environmental factors were analyzed. The results showed that: (1) Compared with the control (CK), the content of soil SOC and TN showed significant changes under the addition of litter; And the C∶N value significantly increased in the treatment of removing the root system of litter (ABR) and adding double litter (AD); (2) The activity of catalase was the highest under the control (CK) treatment, and there was a significant difference compared with other treatments; The activity of polyphenol oxidase was the highest in the treatment of removing litter roots (ABR), and there was a significant difference compared with the treatment of double litter addition (AD) and root removal (BR); In addition, the vector angle of soil enzyme activity was less than 45°, indicating that the soil of the aged Loropetalum chinense forest was mainly limited by nitrogen elements; (3) The results of RDA analysis showed that there was a significant positive correlation between soil extracellular enzymes BX, CBH, NAG content, enzyme C∶N, and enzyme N∶P, and soil available phosphorus and ammonia nitrogen; The content of soil extracellular enzymes LAP, AP, BG, CAT, PPO, and enzyme C∶P were positively correlated with soil available potassium content, but negatively correlated with total phosphorus and nitrate nitrogen content.

In order to clarify the environmental driving factors of soil fertility index of plantation in Guangxi, soil fertility level was comprehensively evaluated. In this paper, the spatial distribution of main soil fertility indexes was analyzed by collecting soil status data of plantations, and the fertility indexes of plantations in the whole region were obtained. Canonical correspondence analysis (CCA) was used to analyze soil fertility indices and environmental factors, and the environmental driving factors of soil fertility indices of plantation forests in Guangxi were obtained. The results showed that the soil fertility index of more than 90% plantations in the whole region was in the range of 0.20-0.50. The order of soil fertility index of different plantations was yellow brown soil > yellow red soil > yellow soil > red soil > limestone soil > lateritic soil > lateritic soil. Soil alkali-hydrolyzable N was highly correlated with organic matter and annual average temperature, soil available P was highly correlated with organic matter and pH, and soil available K was highly correlated with environmental factors including slope, altitude, rainfall, accumulated temperature and slope aspect. Soil alkali-hydrolyzable N was highly correlated with slope and altitude, soil available P was highly correlated with accumulated temperature and rainfall, and soil available K was highly correlated with pH. Soil alkali-hydrolyzable N was highly correlated with slope and organic matter, soil available P was highly correlated with accumulated temperature and rainfall, and soil available K was highly correlated with pH in limestone soil plantation.

Revealing the value of nature reserves in improving landscape pattern in land use can provide accurate information for land use decision-making and regulatory measures. Based on GLC_FCS30-2020 land-use data and a gradient stratification of “Encheng National Nature Reserve- County outside the Reserve” in Daxin County, Guangxi, the moving window method is used to analyze the influence of cropland expansion on karst forest landscape fragmentation. The results indicated that the reserve management was effective, with low forest fragmentation, high vegetation cover, and strong patch connectivity, whereas the fragmentation of cropland landscape was high. The forest cover outside the reserve was appopriate, the fragmentation of forest landscape was high, and the cropland patches were more agglomerated. The cropland expanded to the east, northeast, west and southwest, which resultsed into dramatic changes in the landscape pattern of land used. This study documented that natural reserves could mitigate the threat of the expansion of cropland in karst counties.