C₂₁-steroids is a steroid derivative with 21 carbon atoms in the parent nucleus, and it is a derivative of pregnanes or its isomer. It has many pharmacological activities such as anti-tumor, anti-inflammation and analgesia, anti-fertility, anti-depression and immune regulation, etc. The plant of the family Asclepiadaceae is one of the main sources of natural C₂₁-steroids. According to the statistics of relevant literature, 624 new C₂₁-steroidal compounds were isolated and identified from the medicinal plant of Asclepiadaceae in the past 2 decades, including 9 different skeleton types. In this paper, the distribution, structural characteristics and pharmacological activities of C₂₁-steroids in plants of Asclepiadaceae since 2000s were summarized in three parts, in order to provide references for the further research, development and utilization of C₂₁-steroids, and to provide scientific basis for exploring the medicinal value and therapeutic basis of the plants of Asclepiadaceae. At the beginning of this paper, the distribution, structural characteristics, physicochemical properties and spectra of the prene-type C₂₁-steroids in plants of Asclepiadaceae were reviewed.

The advantages and major problems of platinum-based materials as cathode catalyst for polymer electrolyte membrane fuel cells (PEMFCs) are briefly introduced. The mechanisms of oxygen reduction reaction in acidic and alkaline electrolytes are analyzed and the influences of the adsorption modes of oxygen molecules on the surface of catalyst for the mechanism of oxygen reduction reaction are discussed. The crystalline structures and characteristics of CoSe₂ compound and the relationship with catalytic properties are discussed. The CoSe₂ compound mainly consists of cubic pyrite type (c-CoSe₂) and orthogonal white pyrite type (o-CoSe₂). Generally, c-CoSe₂ has higher catalytic activity than o-CoSe₂. The research progress of the CoSe₂/C catalysts is reviewed. In particular, the effects of preparation method, surface modification, transition metal doping, Co/Se ratio and carbon support on the catalytic activity, reaction mechanism, stability and toxicity resistance of CoSe₂/C catalysts are reviewed. The existing problems and the development prospects of the CoSe₂/C catalyst are pointed out.

A total of 2 047 documents were extracted from the data base of Web of Science Core Collection and subjected to knowledge mapping and visualization analysis with CiteSpace 5.7 R3 software. The results showed that: (1)the whole studied period could be divided into three periods according to the publication numbers (N_P), i.e., the budding exploration period (N_P≤20), the slow growth period (20<N_P≤100) and the rapid growth period (N_P>100), which accounted for about 3.9%, 27.2% and 68.9% of the total publications, respectively. (2)The research between countries was relatively closed, while the cooperation between institutions and authors were loosen. (3)The hot topics varied during the three periods, and the study themes were evolved both in breadth and depth over time. The topics, such as, “heavy metal migration”, “bioconcentration”, “phytotoxicity”, “food security”, and “risk assessment” etc. were the current research frontier topics; and might be continued in the future. The results provided overall frameworks in the field of MnP, and references for further research.

CYP2D enzyme plays an important role in drug metabolism and the detoxification of toxic substances. Knowing and understanding the adaptive evolution of CYP2D subfamily genes helps to understand the mechanism of CYP2D enzymes on toxic substances. Research has shown that CYP2D subfamily genes have lineage-specific expansion in the evolutionary process. In addition, gene duplication, gene conversion, and selective pressure had a certain influence on the evolution of genes, but the effect varied with different lineages. In this paper, the structure, function and evolutionary mechanism of CYP2D subfamily genes were reviewed, in order to provide a theoretical basis for understanding the adaptive evolution of CYP2D subfamily genes.

In order to improve the prediction accuracy of bus travel time, a combined prediction model based on artificial bee colony optimization and extreme learning machine (artificial bee colony-extreme learning machine, ABC-ELM) are proposed after analyzing historical data and the characteristics of traffic flow. First, dynamic and static characteristics like distance between stations, time period and weather conditions are extracted by using IC card and GPS data; after that the dwell time of the station is calculated. Then, the artificial bee colony optimization algorithm (ABC) is embedded in the traditional extreme learning machine algorithm (ELM) to solve the problem of slow convergence speed and difficulty in selecting initial weights and thresholds ELM in bus travel time prediction. Finally, the travel time of the bus on target road section is predicted by using the ABC-ELM algorithm. The model is verified based on the real operating data of Shenzhen Bus 620. The results show that, compared with the widely used BP neural network, SVM and ELM, the method proposed in this paper can maintain lower prediction errors in different road environments and has strong robustness (the RMSE error in peak/off-peak hour is 11.91/8.72, in workday/non-work day is 11.46/9.54,in sunny/rainy day is 10.83/12.31; the coefficient of determination R² in peak/off-peak hour is 0.87/0.92 in workday/non-work day is 0.83/0.88, in sunny/rainy day is 0.89/0.85), which makes it more suitable for travel time prediction in complex urban road environment and for main line bus.

With the continuous improvement of China′s freeway system and the formation of regional freeway network, system network analysis has gradually become an important part of freeway operation and management. As an abstract model of real system, complex network is an important tool to study the network. According to the operation characteristics of freeway system, combined with the road infrastructure network and the traveler travel network, this paper establishes the multilayer complex network model of freeway. On the basis of multilayer complex network, the PageRank algorithm and TOPSIS algorithm are improved, the PageRank-TOPSIS fusion algorithm is constructed, and the importance of freeway nodes is analyzed and evaluated. The results show that the importance of each toll station node varies greatly under different attribute values. PageRank-TOPSIS fusion algorithm can comprehensively reflect the importance of toll station node in complex freeway network according to different weighted regression coefficients. The importance analysis of freeway nodes based on multilayer complex network gives a new perspective to the research of freeway nodes, and provides reliable theoretical support for freeway network analysis and freeway operation management departments.

LS-FIR filter based on least square method is proposed in this paper. According to the relationship between dispersion compensation filter and frequency domain function of the finite length unit impulse response, the linear equations are constructed. Because the number of independent equations is larger than unknowns in the linear equations, the equations have no solution. However, in the sense of minimum norm, a set of solutions is found to minimize the sum of squares estimated error modes. These are the least square solutions, which are used as the tap weight of the filter, and then the filter that needs to be designed is obtained. The simulation results of filter show that under different modulation formats (QPSK, 16QAM, 64QAM) and different signal-to-noise ratios (1~20), in the whole frequency band, when the number of taps is 315, the mean square error reaches 2.610 9×10^-6; In the narrow band, when the number of taps is 197, the mean square error reaches 2.556 6×10^-6; Compared with filters of FIR and Hm-FIR, the algorithm of LS-FIR has higher stability, better approximation effect, and better filtering effect.

In view of the limitations of high energy consumption, difficulty in monitoring and low level of management automation in large-scale greenhouses, the intelligent monitoring system of greenhouse integrating photovoltaic power generation, Low power Wide Area Internet of Things (LPWAN) and cloud platform technology has been designed. According to the climate characteristics of southern Guangxi and the growth characteristics of nightshade vegetables, the light intensity and temperature of the two photovoltaic arrangements in the greenhouse were analyzed by ECOTECT, and it was concluded that the lighting and heat dissipation of the chessboard photovoltaic greenhouse were more suitable for its growth. The greenhouse environmental parameters are collected and transmitted by the LoRa and NB-IoT of the lower computer using its unique advantages in the self-organized star network. Remote monitoring and intelligent supervision are realized on PC and mobile terminal through B/S structure, RBF-PID intelligent control strategy and cloud platform technology. The application shows that the system has a packet loss rate of no more than 3.4% at a communication distance of 1.5 km in a complex environment, and its communication success rate of more than 96%. And it can maintain various environmental parameters within the optimal production threshold for a long time, which helps to improve the output and quality of nightshade vegetables, and also provides a new idea for the intelligent monitoring system of photovoltaic greenhouses.

Non-functional requirements (NFR) describes a set of quality attributes required by the software, such as safety, reliability, performance, etc. In order to develop high-quality software products, it would be beneficial to automatically extract NFR from the Software Requirements Specification (SRS), which not only reduces the labor, time, and mental fatigue involved in identifying specific requirements from a large number of requirements, but also helps developers provide high-quality software that fully meets user expectations. In order to solve this problem, by adopting deep learning feature extraction and classification technology, a BERT-LSTM network model based on the combination of pre-trained BERT word embeddings and long short-term memory network LSTM is proposed, which is used for automatic NFR classification of quality software development. First, use the BERT model to train the word vectors in the sentence. Then use the Long Short-Term Memory network to further perform feature selection and dimensionality reduction. And finally use the Softmax classifier to identify NFR from SRS. Experiments show that in the PROMISE corpus composed of NFR, the BERT-LSTM network model has achieved the best results compared with other algorithms in terms of precision, recall, F₁ score, and other indicators.

The distribution of feature points extracted by traditional fundus image registration methods is too dense, which leads to the inaccurate alignment of the images to be registered images. And the retinal vessel bifurcation feature points have sparse distribution and stable features to improve the accuracy and speed of image registration. Therefore, this paper proposed a deep learning-based fundus image registration framework for vessel segmentation and bifurcation feature point extraction. This framework is composed of two deep convolutional neural networks. The first is the fundus blood vessel segmentation network SR-UNet, which combines channel attention (SE) and residual blocks on the basis of U-Net to segment retinal vessels to assist feature points extraction. The second is the feature point detection network FD-Net, which is used to extract bifurcation feature points from the vessel segmentation map. The proposed registration model is tested on the public fundus registration data set FIRE. The correct matching rate of the feature points is 90.03%. Compared with more advanced retinal image registration algorithms, the algorithm proposed has better performance and strong robustness both in registration quantitative and visual analysis.

The automatic detection of fruits is a key technology in agricultural applications such as automatic picking, orchard spraying, and post-harvesting sorting. Aiming at the problems of small citrus targets, many noises, and serious occlusion in the orchard environment, this paper proposes an improved fast identification method for citrus in the orchard environment based on the YOLOv4 algorithm. The main improvement include: one is to use the Canopy algorithm and K-Means++ algorithm to automatically select the number and size of the priori boxes in the training phase; the other is to add an adjustment layer before each output layer of different scale features in the YOLOv4 network, where the residual network structure is combined with densely connected network, and the loss function of the regression box is modified to detect small citrus in a complex background; third, on the premise of ensuring that a large amount of detection accuracy is not lost, the unimportant channels and networks in the network Layers are pruned. The experimental results of comparison with the three commonly used target detection algorithms show that the improved YOLOv4 detection method in this paper has better detection results for citrus in different growth periods in the orchard environment, with an average accuracy rate of 96.04% and a real-time detection speed of 0.06 s per image, are better than the above three mainstream target detection algorithms. The method proposed in this paper can provide technical and methodological guidance for citrus harvesting and yield evaluation in orchards under natural conditions.

This paper proposes a novel stochastic fluid model (SFM) framework to describe the IP network dynamic video streaming system. The goal is to weigh the relationship between the hysteresis probability and the initial buffer delay, and then to determine an optimal buffer threshold setting. Firstly, the Laplace-Stieltjes transform (LST) matrices of the three First Passage Times (FPTs) is derived in the video streaming system, and then the occurrence probability of the hysteresis is obtained as well as the probability that the initial buffer time is higher than the given tolerable time. Based on these results, the control optimization problem of the initial buffer threshold is further constructed, and an algorithm is designed to find the optimal buffer threshold. Finally, a numerical result is provided to confirm the theoretical findings.

In this paper, the pricing problem on forward-start power options with product of two assets followed jump diffusion models in a stochastic interest rate framework is considered. Using the Feynman-Kac theorem, the joint characteristic function and the Fourier inverse transformation techniques, the closed-explicit solutions of the European forward-start power options with product of two asset are obtained. Some numerical examples for the option price are implemented by the fast Fourier transform (FFT), and the validity of the proposed method is verified by Monte Carlo simulation. The changes of option price in the proposed model were compared with that of other three different models (Black-Scholes, Merton and CIR+Black-Scholes), and the sensitivity of the price of the forward-start power options with product of two assets to some main parameters were analyzed, including the power factor, jump risk factors, maturity date, the correlation coefficient, average recovery speed and long term average level of interest rate, in the proposed model. Numerical results show that the power factor, jump risk factors, maturity date, the correlation coefficient, and long term average level of interest rate have more significant effect on the option price, and average recovery speed of interest rate has some effects on the option price. These are beneficial to risk management and hedging for investors.

Based on the Sprott A chaotic system, by designing a feedback controller, this paper proposes a new four-dimensional autonomous hyperchaotic system with an infinite number of isolated singularities and only two symmetrical singularities. The stability of the non-hyperbolic singularities of this type of new system is discussed by using the central manifold theory, and the Hopf bifurcation is strictly proved. Furthermore, by using numerical methods such as Lyapunov exponential spectrum, bifurcation graph and Poincaré mapping, the existence of periodic attractors, chaotic attractors and hyperchaotic attractors of the new system is obtained.

Habitat height is important for animal to utilize their environment. From July to september 2018, the perch height of crocodile lizard (Shinisaurus crocodilurus) was study by measuring branches in Yusan gully, Chishui gully in Dachai gully in Daguishan Nature Reserve, China. It was found that the large-bodied lizards tend to choose higher perch height, due to the reason that bigger one may needs more food, larger hunting area, more illumination and the ability to avoid enemies or manor competition. In addition, the perch height may be different in different gullies and different months due to different topography, vegetation type and hydrological condition.

In order to explore the expression characteristics of RORα gene in Siniperca chuatsi and the effect of starvation on its rhythm, the expression of RORα in S. chuatsi different embryonic developmental stages and the circadian expression of RORα gene in muscle and liver of S. chuatsi after short-term starvation were detected by real-time quantitative PCR. The results showed that the expression of RORα gene was low in cleavage stage, higher in early blastula to early-gastrula stage, and lower in late-gastrula embryo to larval stage. There was no significant difference from late-gastrula embryo to larval stage except tail bud stage. The expression of RORα gene in muscle and liver of S. chuatsi showed obvious circadian rhythm, which was higher in the daytime and lower in the night; however, the circadian expression of RORα gene in muscle disappeared after 5 days of starvation. Although RORα gene showed circadian rhythm in liver, its median value, oscillation amplitude and peak phase had significant effects (P<0.05). In summary, RORα gene may play an important role in the blastula to early-gastrula stage of S. chuatsi. Furthermore, starvation stress could affect the circadian rhythm of RORα gene in muscle and liver of S. chuatsi, which may lead to the disorder of physiological process.

To explore the potential mechanism of Mahonia fortunei (Lindl.)Fedde and Glycyrrhiza uralensis Fisch decoction for hepatitis based on network pharmacology. TCMSP and TCMID databases are used to collect active ingredients and targets of Mahonia fortunei (Lindl.) Fedde and Glycyrrhiza uralensis Fisch decoction, GeneCards and OMIM databases are used to screen for genes related to hepatitis and to intersect with the target genes of active ingredients in Perilla frutescens and Bidens pilosa L. decoction, a protein mutual network (PPI network) was built by the string website, and the biological process of the intersection target was analyzed by the Reactome database and the KEGG metabolic pathway analysis on the Omicshare platform. A component-target-channel network was built by Cytoscape software, screened core targets and core pathways by conducting network topology analysis method, the core target protein crystal structure was further imported into Autodock software to confirm the molecular docking of the active ingredient and the core target protein. By measuring the biochemical indicators in the acute liver injury model, the mechanism of Mahonia fortunei (Lindl.) Fedde and Glycyrrhiza uralensis Fisch decoction for hepatitis were studied. The result showed that Mahonia fortunei (Lindl.) Fedde and Glycyrrhiza uralensis Fisch decoction contains 95 ingredients, including 872 potential targets. There are 198 targets related to hepatitis in Genecards and OMIM databases. 25 intersection targets were obtained by intersection analysis, and 12 core targets were further obtained from the drug-compound-target network graph. By molecular docking confirmation analysis, three potential lead compounds including Kaempferol, naringenin, and isorhamnetin were screened, and the potential target proteins were confirmed. The experimental results showed that the water extract of Mahonia fortunei (Lindl.) Fedde and Glycyrrhiza uralensis Fisch, kaempferol, naringenin and isorhamnetin can significantly reduce the content of ALT, AST, PNP and MDA in the serum of mice, and increase the content of GSH and SOD. The active ingredients in Mahonia fortunei (Lindl.) Fedde and Glycyrrhiza uralensis Fisch decoction may avoid liver injury by regulation of oxidative stress.

Taking the phase II denitrification operation process of a joint sewage plant in Xinjiang as the research subjects, the response surface method (RSM) was applied to study the influence of sewage influent quality, sludge discharge amount, dosage, sludge reflux ratio, etc. on effluent total nitrogen (TN) and optimize its operating parameters. The experimental results showed that, from the perspective of effluent water quality, the optimal ranges of the influent carbon-nitrogen ratio (C/N), carbon-phosphorus ratio (C/P) and organic load (F/M) were 8.00-9.00, 55.00-60.00 and 0.09-0.10 d^-1, respectively, which is conducive to the nitrification reduction of the system; under the additional factors, the optimal ranges for the amount of sludge, dosage and internal reflux ratio (R) were 0.165-0.170 kg/m³, 14.00-16.00 mg/L and 43.00%-45.00%, respectively, with the best denitrification effect of system. Combined with the mathematical model of nitrogen loss and transfer, it is known that the quality factors of effluent water C/N and F/M had a greater impact on the system load. Therefore, it is helpful to optimize the removal effect of total nitrogen in waste-water by the denitrification unit of the sewage plant when the various influencing factors are kept in the optimal range.