In order to completely identify the competitive market structure of products inter-brands and intra-brands on e-commerce platforms, which contain nearly 100 similar brands and thousands of products of the same brand. Based on the co-occurrence theory, a study is conducted on the brands and products appearing in the same browsing behavior of online consumers as competing relationships from the online browsing behavior of consumers, and social networks are used to visualize inter-brand and intra-brand competition from the perspective of consumer browsing behavior. The analysis results show that the homogenization competition of products intra-brands is more serious than that of inter-brands. This result has important practical significance for the production line management of products within the brand. Compared with focusing on the competitive relationship between inter-brands, brand manufacturers should put more time and energy into the design of the length of production line of intra-brands, so as to rapidly enhance the brand's market competitiveness.

Dimensionality reduction becomes a significant problem due to the proliferation of high dimensional data. For dimensionality reduction, feature selection is more analytical than feature extraction. Therefore feature selection plays an important role in pattern recognition. In recent years, many feature selection methods based on manifold learning have been proposed. However, such methods are susceptible to noise data. Therefore, an efficient unsupervised feature selection method is proposed-joint low-rank representation and graph embedding for unsupervised feature selection (JLRRGE). This method not only finds the low rank structure of the data after selecting feature, which makes the algorithm more robust, but also preserves the local manifold structure of the data through the adaptive graph embedding. The experimental results show that the proposed algorithm is superior to other compared methods.

The causal relationships between variables from dataset and the corresponding causal network can be recovered by the constraint-based methods. However, in high-dimensional dataset, the accuracy and efficiency of the constraint-based methods are not high, which seriously affects the application and promotion of such methods in high-dimensional dataset. In order to solve these problems, a causal network structure learning method based on low-order conditional independent (CI) test is proposed. In this method, CI test based on lower order conditional set is used to construct a rough causality skeleton; and the split-merge strategy is taken to divide the large rough causality skeleton into a set of smaller subnetworks. The structure of each subnetwork of lower dimensionality is constructed independently, thus able to improve its accuracy. Finally, a complete causality network graph is integrated. The experimental results demonstrate the technical feasibility.

A metasurface broadband Terahertz (THz) absorber is presented, of which the unit cell is made up of a split elliptical irregular ring, a dielectric substrate and a metallic ground. The elliptical irregular ring is composed of a smaller ellipse eroded from a bigger ellipse, and the two ellipses have different ratio of the long axis and the short axis. The simulation results show that the absorber achieves a broadband absorption from 0.98 THz to 2.64 THz, with the absorptivity over 90% at normal incidence, the maximum absorptivity being 99.99%, which are in agreement with the calculated result by using an interference model. The bandwidth is 91.7% with respect to the central frequency. Besides, the proposed absorber has property of wide angle absorptivity for both TE and TM polarizations. Compared with published designs, the proposed THz absorber has a simple geometry but a broadband and polarization insensitivity, and hence it can be used in many applications, such as THz imaging system, radiometer and stealthy technology.

Aiming to study human anticipatory postural adjustments that respond to the external interference impacts, and enabling the possibility of theoretical guidance for human-robot interaction, ten male students are selected to participate in this study and complete human-robot interaction experiments with heavy shock interferences, responding without visual and auditory information. And the heavy object posture information and the SEMG signal of the human arm are collected synchronously to calculate the APAS activation time and heavy object deflection degree of the experimenters' muscle. The APAS activation time is considerably greater in visual information and auditory information (p<0.001). APAS not only guarantees the heavy object deflection degree decreases significantly, but may also be important in the stable motion collaborative posture control task. The visual information and auditory information lead to premature muscle APAS and make motion posture control more stable and safer in the mechanism of collaborative tasks.

A novel Parallel Hybrid Active Power Filter topology without transformer proposed is against the problem of excessive subharmonic content and insufficient reactive power in some cases, which consists of active and passive parts, and it increases active part to compensate harmonics on the basis of passive part to realize reactive power compensation. The system makes use of the working characteristics of the passive part series partial voltage and series resonance to reduce the active part working voltage low, and then reduce the capacity of the active part. After analyzing the working characteristics of the new topological circuit, a mathematical model of the system is established, using the decoupling control method to achieve effective control of the system. Experimental results confirm the feasibility and advantages of the proposed method.

The metabolic model of colon cancer cells is studied. The model contains five coupled reaction diffusion equations, in which some equations involve discontinuous terms. It is proved that this problem has a global solution by using the L^p-theory for parabolic equations, the Schauder Fixed Point Theorem and approximation method.

In this paper, the common fixed point theorem for mapping pairs under c-distance is obtained in the cone metric space, The results do not require nondecreasing in the condition. In addition, the condition of the results neither require the normality for cone nor require the continuity for mapping, Furthermore, the existence and uniqueness of the mapping to the common fixed point under c-distance is proved. The results improve and generalize the relevant conclusions in the existing literature.

"Liquid-separation condensation" theory is applied for performance enhancement in the plate condenser, with the physical and mathematical models built. Seven liquid-separation plate condensers having the same heat transfer area are proposed for heat transfer coefficient and pressure drop comparison. Their performances are evaluated by employing the penalty factor (PF) and exergy destruction. Results show that the higher the liquid-vapor separation efficiency, the better is the performance of the liquid-separation plate condensers. Compared with the conventional plate condenser with the same heat transfer area, the liquid-separation plate condenser has 14.6% lower PF and 2.6~6.1% smaller exergy destruction when the liquid-vapor separation efficiency is 100% and the mass flux is in the range of 0.08~0.12 kg/s. Moreover, when the mass flux is 0.1 kg/s, the penalty factor (PF) of liquid-separation plate condenser is lower with the vapor quality ranging between 0.3~0.75, and the exergy destruction of liquid-separation plate condenser is superior only when the vapor quality is higher than 0.45.

Based on the pavement technical condition of the test highway S120, a three-dimensional model of the thicker asphalt layers overlaying on the existing concrete cement pavement (CCP) is established by the finite element software ABAQUS, which considers the interlayer contact between the existing CCP and overlay layers and the joints in CCP. First, the temperature field of the structure is calculated and analyzed, and then the influence of the thickness of adjustment layer on the mechanical response of the structure under the coupling action of temperature and traffic load is studied. At last the functional requirements of each structural layer are analyzed. The following conclusions are drawn:under the local typical high temperature, the temperature in the top surface is higher than 55℃, and the temperature field varies greatly between 0~30 cm; increasing the thickness of the overlay layers reduces the temperature stress caused by the existing CCP. The overlay layers are subjected to larger vertical displacement and the upper surface layer and adjustment layer are subjected to larger tensile stress, which is the main causes of Surface-down cracking and Bottom-up fatigue cracking. The upper surface layer is subjected to large compressive and shear stresses, while the lower surface layer and adjustment layer are also subjected to large shear stresses; as the thickness of adjustment layer increases, the tensile, compressive and shear stress of adjustment layer can be effectively reduced, and the required tensile strength of the old CCP structure can be greatly reduced, which enhances the anti-reflective crack ability and anti-fatigue ability of structure; according to the mechanical analysis, the proposals of the Mix Designing of the overlaying layers are given. The research results providing a theoretical basis for improving the design of road materials and improving the durability of overlay structures.

The effects of high content Al₂O₃ particles on the apparent activation energy, thermal decomposition kinetics and properties of epoxy resin (E51)/4, 4'-Diaminodiphenylmet (DDM) system were investigated by a thermal analysis. The results on the curing kinetics of non-isothermal differential scanning calorimetry (DSC) showed that the addition of Al₂O₃ promoted the curing reaction. The activation energy of the E44/DDM system was reduced from 51.49 kJ/mol to 48.12 kJ/mol after 80%Al₂O₃ addition. In addition, the kinetic parameters of the curing reaction were obtained by the n order model. Thermo-gravimetric analyzer was used to study the thermal decomposition kinetics of the epoxy curing system. The FWO method was used to calculate the activation energy of the degradation. The results showed that the addition of Al₂O₃ powder had little effect on the initial decomposition activation energy of the 80%E51/DDM system. When the degradation rate reached 30%, the Al₂O₃ powder had obvious inhibitory effect on the decomposition of E51/DDM system. The results of TG-IR test showed that carbon oxides, methane, carbonyl compounds, amines and aromatic compounds are the main products in the process of E51/DDM thermal decomposition. Al₂O₃ powder can improve the thermal stability of the E51/DDM system. Dynamic mechanics showed that the addition of Al₂O₃ increased the modulus of storage of epoxy resin cured products. DSC results showed that the addition of Al₂O₃ powder increases the glass transition temperature from 114.16℃ to 121.51℃ of the system.

Photoisomerized energy storage materials of azobenzene which can capture, convert, store, and release solar energy by reversible photoisomerization have become one of the main focuses of current research. In recent years, certain research results have been achieved using azobenzene for light energy storage. Based on the energy storage mechanism and main performance parameters of photoisomerized energy storage materials of azobenzene, the progress of half life, energy density and visible light energy storage of photoisomerized energy storage materials of azobenzene are focused on, and the merits and drawbacks of those azobenzenes analyzed. Finally, a prediction is made concerning the research direction and application prospect of photoisomerized energy storage materials of azobenzene in solid film.

The design method and its working principle of a sliding linear Fresnel solar collector are presented. When the true sun was from 10:00 to 14:00, the optical path simulation of the collector was performed in TracePro, so the change of the spot in one day was obtained and compared with the experimental results, and finally the annual spot simulation analysis was performed. The simulation and experimental results show that the two curves are basically the same, and they both show the parabolic symmetry law of increasing first and then decreasing. The average collector efficiency of one day is 22.90%, and the simulated equivalent average heat collection efficiency 24.45%, the average heat collection efficiency without slip 6.40%.

It is an urgent problem to further promote the reform of science and technology system and increase the ability of innovators to promote innovation-driven development. As an important part of our modern scientific research organization system, new R&D (research and develpment) institutions are not only a breakthrough in deepening the reform of scientific and technological system, but also an important way for the transformation of national innovation power. Considering the characteristics of the times and the key influencing factors, an endeavor is made in dividing the development of new R&D institutions into four stages, and concisely summing up the development characteristics of new R&D institutions from the perspectives of policy design and practice exploration at the national and local levels, aiming at providing a panoramic perspective for further understanding the growth path of new R&D institutions as a new scientific research organization.

The new R&D (research and development) institutions are not only the new force and bridgehead for the construction of national innovation system, but also the "special zones" for the exploration and establishment of modern scientific research institute system and the most dynamic "experimental fields" for the reform of national science and technology system. Starting from the process of reform of science and technology system, the policy design and practice exploration of national and local science and technology innovation is reviewed. Focusing on the objective needs of deepening the reform of science and technology system and the inevitable choice of the transformation of national innovation power, a discussion is conducted on the conceptual connotation of new R&D institutions as "experimental fields" and "special zones" of China's science and technology system reform, as well as the new type of R&D institutions. The background and realistic opportunities of R&D institutions are systematically discussed. The purpose is to provide theoretical support for the formulation of relevant policies, guidance and support, as well as to promote the high-quality development of new R&D institutions, and lay a foundation for further analysis of the formation motivation, strategic positioning and function of new R&D institutions in China.