Shopping behaviors in the social network can reflect users’ willingness to meet their material needs and social needs under the influence of social interaction, which is an important research in social network marketing. The traditional analysis of online shopping behavior only focuses on the similarity between users’ behaviors while ignoring the influence of users’ social needs and peer behaviors. For that, the features of users’ shopping behavior are constructed by combining anti-conformity theory and social needs. Secondly, aiming at the incomplete observation of user data in social network, a causal mechanism discovery algorithm for users’ behaviors based on Fast Causal Inference (FCI) is proposed. Finally, the rationality of the causal mechanism of our model is verified based on the experimental analysis and empirical analysis.

Intrusion detection is very important for network security. Traditional intrusion detection algorithms are often affected by biased training samples and misleading characteristics of attack behaviors. Therefore, a self-correction small sample classifier for intrusion detection is proposed. First, an orthogonal projection classification method roughly divides training data set into three groups. Then, based on support vector machine and random forest algorithm, sub-classifiers are constructed layer by layer to refine the results iteratively. Finally, by combining results of all sub-classifiers, a classifier for the NSL-KDD data set is constructed. Experimental results show that the proposed classifier surpasses its competitors in the detection accuracy of DoS (Denial of Service), Probe and R2L (Remote to Local). The overall recall and accuracy rates are better than others.

Aiming at the problems of slow speed, subjectivity, high requirement of professional knowledge and high recognition cost in the identification methods of crop pests and diseases in China, a crop pest and disease identification method is proposed based on Inception-resnet-v2 network model combining Inception and Residual structure. To achieve accurate and efficient crop pest identification, the network uses a residual structure to combine low-level features with high-level features to increase network depth. At the same time, the Inception structure is added, which not only maintains the sparsity of the network structure, but also utilizes the high computational performance of the dense matrix to speed up the training. Finally, the multi-class prediction is performed by the Softmax classifier. Compared with the traditional method, the proposed method converges faster, not only with an accuracy rate of 96.67%, but also with an accuracy of 90.77% and a recall rate of 89.72%. It also makes the different categories of pest and disease identification more balanced, improving the specific methods in the traditional methods, by which the recognition effect is poor.

Approximate Nearest Neighbor (ANN) search has served as one of the most important technologies for efficient retrieval of large-scale data in the era of artificial intelligence. As a promising solution to the ANN, hashing has received a lot of attention due to its high efficiency and extensive works have been presented in the literature. However, so far, there is no work with attempt to comprehensively analyze and overview the state-of-the-art hashing methods. To address this, the basics of hashing, including distance calculation, loss function, discrete constraint and out-of-sample learning, are first systematically introduced. Then, the state-of-the-art hashing based methods are comparatively studied and experiments on the widely used databases are conducted to evaluate their performance. Finally, the key problems of hashing methods are summarized and some potential research directions are pointed out. It is believed that this endeavor could provide other researches with a useful guideline in designing effective and efficient hashing methods.

Network centrality is important for network analysis. Therefore, n-Degree and k-Stress centrality are proposed to enrich the analysis options using network centrality. On the other hand, for large scale networks, parallel algorithms are necessary for fast computation. To this end, parallel algorithms for the n-Degree and k-Stress centrality are designed and implemented along with 11 existing centralities in a toolkit. The toolkit is based on Spark Pregel, and tested on the BoardEx social network. Experiments show the feasibility and scalability of the toolkit.

The safety of high-voltage power line is usually threatened seriously by the foreign matters such as bird’s nests or kites and so on. The transmission tower is an important part of the system of high-voltage power line. So the UAV (Unmanned Aerial Vehicle) often takes photos of tower especially when inspecting the high-voltage power line and these photos usually have to be analyzed by the classical algorithm of YOLOv3 (You Only Look Once Version 3) whether they contain the foreign matters or not. This research is conducted to improve the classical algorithm of YOLOv3 in order to improve its precision and deficiency and unknown scale of weight parameters. Initially, the structure of improvement is designed and image data set is constructed. Second, the classical algorithm of YOLOv3 is improved from three ways: the width and height of loss functions of prediction box, the unbalanced loss function of prediction type and the network structure of classical algorithm are improved respectively. The experiments show that: the improvements proposed are effective, which improves the average recognition precision and reduces the scale of weight parameters greatly and maintains the efficiency. The result of improving the classical algorithm’s neural network structure is obviously better than other ways, which is probably the main direction to improving the algorithm. The investigation provides important basics to detect the objects on real time of UAV.

In the virtual surgery system, soft tissue deformation simulation will directly affect the reality of the simulation. In this research, a physical model of gingival tissue involved in oral surgery is established by means of mass-spring model. A midpoint-averaging algorithm is proposed for the simulation of gingival deformation in virtual surgery, which combines the advantages of the midpoint method and the average method, and uses the midpoint method to solve the speed and the average method to solve the displacement. It can effectively reduce the error and reduce the impact of step size. Compared with a variety of Explicit Euler algorithms, it is further verified that the midpoint-averaging algorithm has a higher solution precision and is less affected by the step size.

When the motion control algorithm is implemented on the FPGA(Field Programmable Gate Array), and the data type changed from floating point to fixed point, there is a problem that high precision cannot be guaranteed. To solve this problem, the range of each signal data in the algorithm is analyzed using the Fixed-Point Tool in simulink, and then the suggested bit width given by the Fixed-Point Tool is artificially modified. Therefore, the algorithm can solve the problem of resource waste caused by using a unified longer bit width on the FPGA while ensuring higher precision. Compared with the traditional artificially defined data bit width, the Fixed-Point Tool is used to set the data bit width, and the internal data bit width of the algorithm can be more flexibly and dynamically adjusted when the input data range changes.

As an extension of topological entropy, topological pressure was first proposed by Ruelle in 1973, and then further extended by Walters to compact metric spaces of continuous maps. It reflects the complexity of dynamic system, and it is an important research content in fractal geometry and dynamic system. In order to solve more complex problems relevant to topological pressure, more and more new research has emerged. On the basis of the topological entropies defined by Biś and Patrão, the topological pressure of a semigroup action generated by proper maps of the metric space is given, and then the topological pressure of a free semigroup action generated by proper maps in a locally compact separable metric space is proved, which is equivalent to the topological pressure of a free semigroup action in its one-point compactification space. Furthermore, some properties of the topological pressure of a free semigroup action are given.

A numerical simulation method of multiple-beam interference to produce three-dimensional photonic lattice structure based on Matlab GUI was proposed. The photonic lattice structures could be observed intuitively by calculating numerically the light intensity distributions of the interference patterns. Moreover, the influence of some parameters on the structures was discussed. In the concrete simulation process, (6+1) beams to interference, where the central beam is a vortex beam, was used, and the remaining six beams distributed symmetrically around the center beam. In the past, the method to study the photonic lattice by experimental device was with serious errors, involving complex device adjustment and long experiment period. Nevertheless, the numerical simulation method based on Matlab GUI in the study has not only high efficiency, low error and intuitive results, but also flexible and fast parameter adjustment, user interface friendliness, and other advantages.

The wettability of the solid surface is affected by surface roughness and surface microstructure. On the basis of analyzing and summarizing several common surface microstructure processing techniques, the jet mask electrolytic processing method is proposed to process the micro-pit array structure on the surface of stainless steel, and study the influence of processing voltage on the size of the micro-pit and influences of the wettability on the microstructure. The stainless steel micro-pit structure is used as a mold, and the micro-convex structure is injection molded by PDMS (polydimethylsiloxane). The static contact angle is measured by fluorosilane and the contact angle is (151.7±0.51)°.

With the emergence of a large number of distributed energy stations and the popularity of electric vehicles, the security and economic operation of the power system are affected. At the same time, the traditional load forecasting methods are also facing challenges. To solve this problem, a Whale Optimization Algorithm-Least Squares Support Vector Machine (WOA-LSSVM) is proposed for short-term power system load forecasting. Using the advantages of whale algorithm such as strong global optimization ability and fast convergence speed, the blindness of parameter selection of LS-SVM is overcome, and the load forecasting accuracy of LS-SVM is improved. WOA-LSSVM is used to forecast the load of a certain area in Zhejiang province in the next day based on the historical load data in 2013. The results are compared with those of the particle swarm optimization least squares support vector machine model and the standard LSSVM model. The results show that the short-term load forecasting based on the whale optimization LSSVM has a higher forecasting accuracy and a smaller relative error.

Cooking in the residential kitchen is an important source of indoor PM_2.5. In order to control PM_2.5 effectively, the air curtain is proposed as an air supply means. A model of the kitchen is built and a numerical simulation, including airflow organization, temperature distribution and PM_2.5 concentration distribution in the kitchen, is performed by using the Fluent software. A numerical simulation is performed to study the control effect of PM_2.5 pollutants and heat flow in the kitchen. Three kinds of jet velocities are compared and analyzed. The results of this study show that the jet flow of an air curtain has a good wrapping effect on the cooking area, which can obstruct the diffusion of PM_2.5 and the spread of heat flow. The PM_2.5 removal rate can be increased to 44%~75%, and the average temperature in the kitchen can be lowered by 1~2 ℃ by using the air curtain to supply air. The control effect is optimal when the air curtain jet velocity is 0.6 m/s. The research productions have reference for the prevention and treatment of PM_2.5 pollution in the kitchen and can provide simulation data and a theoretical basis for the study of the air curtain air supply system.

Dihydromyricetin (DMY) is an important active substance in rattan tea, with antioxidant, antibacterial, anti-tumor, anti-fatigue, anti-inflammatory, anti-alcohol, liver protection and other bioactive effects, which is widely used in the food and pharmaceutical industry. In this study, the structure identification, content determination, efficacy and pharmacokinetics of DMY were reviewed, so as to provide theoretical basis for the in-depth research and further development and utilization of DMY.

Small molecular inhibitor with Glu-Urea skeleton of Prostate Specific Membrane Antigen was synthesized by 7 steps using L-glutamic acid and L-lysine as starting materials. The yield of each step was higher than 60% except for the synthesis of L-di-tert-butyl glutamate. The total yield of the route was 18%. The products were characterized by ¹H NMR, ¹³C NMR and HR-MS(ESI). A new synthetic method was designed for the synthesis of intermediate tert-butyl N⁶-(((9H-fluoren-9-yl)methoxy)carbonyl)-L-lysinate, which realized the deprotection of lysine alpha amino group and the protection of carboxyl group by Boc in one step, simplified the reaction steps and increased the synthetic yield of compound 3 from 20% to 70%.

The outstanding corrosion resistance performance of FRP bars can avoid the deterioration of concrete structure which is mainly caused by the bar corrosion. Therefore, concrete reinforced with FRP bars has attracted the attention of researchers at home and abroad. The bond behavior between FRP bars and concrete and the anchorage length of FRP bars are the main parameters which can influence the cooperative work of FRP bars and concrete. In this research, pull-out tests of different FRP bars embedded in concrete specimens are carried out. The damage mechanism and bond strength are analyzed. And the formula to estimate the anchorage length of FRP bars is derived, providing the theoretical guidance and basis for the practical engineering.

In order to analyze the back electrode contact characteristics of CZTSSe thin film solar cells, the CZTSSe/Mo(S,Se)₂/Mo structure was simulated by AFORS-HET v2.5 software. The effects of the band gap E_g and the electron affinity of CZTSSe and the thickness and the band gap of Mo(S,Se)₂ interface layer on the electrical contact characteristics between CZTSSe and Mo electrodes were investigated. The results show that the increases of the band gap and electron affinity of CZTSSe make the ohmic contact of CZTSSe/Mo(S,Se)₂/Mo weaken and transform to the rectifying contact. For CZTSSe with a narrow band gap, the addition of the interface layer causes the ohmic contact formed by CZTSSe/Mo(S,Se)₂/Mo to be converted into a rectifying contact. As the thickness of the interface layer increases, the rectifying contact weakens. For CZTSSe with a wide band gap, the addition of a 2 nm interface layer enhances the rectifying contact formed by CZTSSe/Mo(S,Se)₂/Mo. As the thickness of the interface layer increases, the rectifying contact weakens. When the work function of CZTSSe is low, CZTSSe/Mo(S,Se)₂/Mo forms an ohmic contact, and the optimal electrical contact characteristics can be obtained by controlling the thickness of the interface layer to about 100 nm.

The temperature slip characteristic of zeotropic-mixtures, which can effectively improve the performance of the ORC system, has an essential investigation value. First the ORC system test bench was introduced, and then a zeotropic-mixture system model for a thermodynamic analysis was established. Next, the effect of zeotropic-mixture mass flow rate on the ORC system operation performance and the heat transfer characteristics were experimentally studied, and the results compared with those of the ORC with pure working medium system. The results indicate that the performance of the ORC utilizing zeotropic-mixture R1234ze (E)/R245fa has an obvious improvement.