Abstract:

Abstract: Abstract: In order to increase the risk factors that could be accepted and reduce the data format requirements in cardiovascular disease (CVD) predicition models,machine learning algorithms were used to change the strict mathematical formulas of traditional CVD prediction models.Frstly,a CVD prediction model by extreme learning machine (ELM)algorithm based on single hidden layer feed-forward neural network (SLFNs) was proposed. Further more,an enhanced leader particle swarm optimization(ELPSO) through a five-staged successive mutation method was used ,and the optimized strategy of PSO was also used to optimize the SLFNs hidden layer units parameters.The analysis results on Statlog(Heart) dataset and Heart Disease Dataset of UCI database indicated that the test accuracy of proposed ELPSO-ELM model could reach 85.71%and 84.00%respectively, the AUC(The area under the ROC curve )could reach 0.9024 and 0.8423 respectively.They were higher than conventional CVD prediction models.The proposed models relaxed the linear constraints of data format and more complex risk factors could be accepted.

Abstract: As the transmission lines are located in the wilderness where the thunder is continual, lightning waves intrude into substations along the lines, causing frequent damage to the electrical equipment inside the station. In this paper, taking the lightning accident of a substation as an example, the ATP-EMTP electromagnetic transient software builds the whole model containing the line section and the substation. Then it calculates the lightning overvoltage and waveform of the main transformer and other electrical equipment. We analyze the main causes of the lightning accident and study how the lightning strike point and the grounding resistance of the pole and tower influence the station. Finally, we propose some corresponding protective measures. The results show that the accidents are related to the overshot of lightning current invading wave and the long distance between lightning arrester and transformer, and installing lightning arrester in front of the transformer can effectively avoid such accidents

Abstract: : In order to solve the problem that the IPv4 network architecture and equipment are difficult to meet the scale deployment and application of the IPv6 network in terms of packet parsing, matching, and action execution. A new data plane structure supporting IPv6 experimentation and deployment is proposed. The structure includes a data plane abstraction mechanism that supports both policies and functions and a data plane hardware structure applied to the IPv6 Next Generation Internet. It supports a variety of innovative network architectures coexisting in the same network, enabling the testing and verification of new protocols. It supports customizable protocol resolution, flexible programmable packet processing, and dynamic composition of internal resources. These are used to support the testing, deployment, and evaluation of future network functions. Through system experiments and analysis, it is verified that the structure has high forwarding performance under the condition of acceptable resource .

Abstract:   In order to solove the increasingly serious congestion problem of urban roads ,a road congestion dectection method based on convolutional neural network was proposed in this paper.This method was based on the appliciation value of road congestion rapid detection technology to alleviate the traffic jam promblem.Besides,it combined deep learning with image processing  technology.Compared with traditional methods,this method did not need to extract the background in the early stage, and was not affected by the illumination brightness and the actual environment.It had the characteristics of fasrt recognition speed,less ocupied comoutong resources and good generalization. It has been applied in practical projects and achieved good results.

Abstract: Since PSO algorithm was casy to get trapped into local optimum ,in this paper ,based on the learning theory a new PSO algorithm nameed as L-PSO was proposed. In L-PSO,an inger value was set as the maximum cycle limitation for the global best particles ,and propose a clustering grouping mutation mechanism which could devide the particles into some subb-swarms and generates the competitive by crossover and mutation of the teo centers selected randomly from sub-swarms.Then the competitive particle was used to replace the global optimum particle which could help jump oout of the local optimum and improve theconvergence speed. Experimenta results on several benchmark test function shoxed that L-PSO was very effective. 

Abstract: Since PSO algorithm was easy to get trapped into local optimum,in this paper, based on the learning theory a mew PSO algorithm named as L-PSO was proposed. In L-PSO ,an integer value was set as the maximum cycle limitation for the global best particles,and propose a clustering grouping mutation mechanusm which could devude the particles into some sub-swarms,Then the competitive particle was used to replace the global optimum particle which could help jump out of the local optimum and improve the convergence speed. Experimental results on several benchmark test functions showed that L-PSO was very effective.

Abstract: The traditional extended Kalman filter algorithm has been widely used in permanent magnet synchronous motor position sensorless control system. This method has good dynamic performance, but it still has problems in estimating the accuracy of rotor speed and position, especially During motor startup and low speed operation, due to the small induced potential, the extended Kalman algorithm has large deviations in the estimation of rotor speed and position. In order to solve this problem, this paper designs a comp osite detection method that combines the extended Kalman filter algorithm with high-frequency signal injection method. The Kalman filter method is used in medium and high speed, and the high frequency signal injection method is used in zero speed and low speed. The transition area is designed with a special algorithm that fuses the two methods to achieve full sensorless high-precision control in the full-speed domain. Finally, the simulation results show that the permanent magnet synchronous motor sensorless vector control system based on this hybrid detection algorithm has the advantages of fast dynamic adjustment, good robustness and so on.

Abstract: A novel filtering method based on the extended state observer (ESO) is proposed according to the engineering practice that the system outputs are impact by measurement noise. The linear ESO (LESO) filter achieves noise reduction process based on the accurate observation ability of ESO. And the LESO filter doesn’t need the statistical characteristics of noise in its application and it is capable for control system outputs with unknown or varied sampling rate. The LESO filter is simple in structure with two tunable parameters and is suitable for embedded systems. Proof of the uniform convergence of the LESO for control systems with output measurement noise is presented in the analysis section. And the LESO filter is presented in discrete form. Simulation results reveals that the linear ESO filter outperforms the Kalman filter with less computing time and higher precision. Then the proposed filter is applied to the wind identification of the powered parafoil and payload system in landing area. The application shows that the LESO filter is effective and easy to use

Abstract: Integrating user’s social information is an appropriate way to solve the user-cold start problem. Though various prediction models focus on integrating social relation information, few have noticed the dynamic change of the user’s interest. Thus, in this paper, we propose a cross-domain temporal interest prediction approach by integrating social activity information. First, we construct a cross-domain personized ranking model which can map the feature from social space into the purchase space. Further, we propose a feature modeling method based on data grouped by time period. Experiments on the dataset verify that the proposed method can predict user’s interest more effectively.

Abstract: The traditional hospital division is too professional and medical treatment process separates. In the absence of good consultation service, patients are very inconvenient. With the development of information technology, more and more hospitals begin the construction of consultation system, but the function of these systems is still not comprehensive. The consultation system in this paper is used for 3D modeling of the Chinese Traditional Medicine Hospital of Xiamen, with walking, roaming, automatic navigation, hospital department introduction and other functions. The system is based on Windows platform, using 3DMAX, MAYA to build 3D models, and interaction is implemented in Unity3d. The core of interaction is automatic navigation. It use A* algorithm. According to the relative fixation of the objects in the hospital and the crisscrossing of the road, the heuristic function of Manhattan is adopted to simplify the algorithm. The system can effectively help patients understand the medical environment, and the operation of the system is simple and intuitive, which is convenient for patients among different ranks.

Abstract: Polyethylenimine@Fe3O4 magnetic composites (PEI@Fe3O4) were prepared by a surface overlay method and characterized. Its adsorption property toward Congo red (CR) from solution was performed in batch mode. The results showed that it was favor of CR adsorption onto PEI@Fe3O4 with the increase of ionic strength. Adsorption isotherms from experimental results were fitted well by Redlich-Peterson model and the adsorption quantity from experiment was up to 58.3 mg/g at 303 K. The kinetic process was best predicted by the Elovich model. Desorption and regeneration experiments showed that the adsorbent can retain its adsorption capacity even after adsorption/desorption cycles, showing excellent mechanical and chemical stabilities. It was implied that PEI@Fe3O4 be promising in removal of anionic dyes from wastewater

Abstract: Cellulose,nad loose pore and long chain structure are favorable for the permeation and adsorption of macromolecules. When the calcium chloride solution was added to the mixed solution of cellulose and sodium alginate, cellulose and alginate can be cross-linked with calcium ions to form the gel. So glucoamylase could be immobilized by the carrier materials mixed with cellulose and sodium alginate. Some factors which affect the immobilized enzyme activity were considered, such as mix proportion for cellulose and sodium alginate, mass concentration of the mixed solution for cellulose and sodium alginate, and mass concentration of calcium chloride solution. The response surface methodology was carried out based on single factor test, and the optimal conditions for preparing immobilized glucoamylase were obtained as follows: the ratio of cellulose and sodium alginate was 1.5:1, the mass concentration of the mixed solution for cellulose and sodium alginate was 2.80%, the mass concentration of calcium chloride solution was 2.77%. The optimal conditions’ result of response surface methodology was tested successfully. In addition, the reusability test was conducted, and the results showed that the immobilized glucoamylase which adsorbed to the enzyme solution was better than unadsorbed and immobilized glucoamylase.

Abstract: In this paper, the graphene (G) /thermoplastic polyurethane (TPU) /polydimethylsiloxane (PDMS) conductive polymer composites(CPCs) with conductive fibrous network s were prepared by the electrostatic spinning and spraying technology, and their strain sensing behaviors were studied in detail. The CPCs possess large elongation at break (165%). With the strain increasing , the sensitivity increases gently at first , and when the strain is greater than 50%, the sensitivity increases sharply, which shows a positive response between strain and sensitivity. Meanwhile, the CPCs exhibit weak hysteresis effect and stable resistance-strain sensing response behavior during the cyclic stretching, which is attributed to the encapsulation of PDMS . For the cyclic resistance-strain sensing curve s, the CPCs show good response stability under different strain amplitudes. The G/TPU/PDMS CPCs reported in this paper have brilliant application prospects in the field of flexible wearable electronics.

Abstract:  A kind of PAN-based weak alkaline ion exchange fiber was synthesized and was used as adsorbent for Vanadium. The adsorption and regeneration performance of two different types of functional groups of ion exchange materials were compared. Effects of pH, temperature, time and the concentration of vanadium on fiber saturated adsorption amount were investigated. The optimal concentration of sodium hydroxide for desorption the fiber was discussed. The results show that the static adsorption capacity of Cl? form fiber are more superior; The maximum adsorption capacity (294 mg?L?1) is obtained under pH 6~8 and adsorption capacity increases with the increase of temperature, the adsorptive equilibrium can be achieved with 6 hours, and the removal ratio for Vanadium was more than 98%. The desorption rate of the fiber is above 98.02%, when the concentration of desorption solution (sodium hydroxide) is 0.5 mol·L－1.

Abstract: Sodium tantalum oxide (NaTaO3) powders were synthesized using tantalum oxide (Ta2O5) and sodium carbonate as precursors, NaCl-KCl-NaF and NaCl-KCl as molten salt medium. Effects of the types and amount of molten salt, reaction temperature on the phase composition and microstructure of the as-prepared samples were studied; they were characterized by X-ray diffraction and scanning electron microscope, respectively. When the ratio of molten salt and reactant precursor is 1:1, only NaTaO3 phase can be detected in the samples after firing at 600 ℃ in NaCl-KCl-NaF molten salt medium. On increasing the reaction temperature to 1000 ℃, the particle size of as-prepared products is less than 1μm, and their morphology of NaTaO3 is cubic-like shape. The complete convention temperature of precursors to NaTaO3 is 700 ℃ in NaCl-KCl molten salt medium, and the cubic-like shape NaTaO3 powders are fully synthesized at 900 ℃. The as-prepared NaTaO3 powder possesses good photocatalytic properties, and the photocatalytic property is correlated to the particle size and morphology of NaTaO3 powders.

Abstract: Impact effects on carbon emissions intensity by population, per capita GDP, and main types of energy in China were evaluated with the fixed effect model based on LSDV estimation with reasons of the results of Likelihood Ratio Test and Hausman Test. The traditional model of STIRPAT was improved by adding Carbon Emission Intensity and Energy Consumption Variables, which included consumptions of coal, coke, crude oil, gasoline, kerosene, diesel oil, fuel oil, and natural gas, except population and per capita GDP. The results show that consumptions of different types of energy have different impacts on carbon emissions intensity from 2004 to 2016 in China. Five variables of energy consumption, which were corresponding to coal, coke, gasoline, diesel oil, and natural gas, had played positive effects on carbon emission intensity from the data of China Statistical Yearbook and China Energy Statistical Yearbook of 200 5 to 201 7. Other variables of crude oil consumption, fuel oil consumption, and kerosene consumption took opposite impact on carbon emission intensity. Moreover, change of population had the most significant favorable influence on carbon emission intensity in all studied variables. Unfortunately, per capita GDP and coal consumption contributed to the increasing of carbon emission intensity in China in the studied period.

Abstract: The technology of high field asymmetric ion mobility spectrometry (FAIMS) with both huge resea rch value and broad application, has great advantages in the field of chemical detection. The principles of FAIMS technology were briefly introduced in this paper, as well as the current research update and application s. In this study, a set of self-designed apparatus was built and used for the determination of a series of FAIMS parameters under different humidity conditions. Within the experimental range of humidity, the peak values were linear to the humidity when DF = 50 % ; the signals of PIP were continuous in 3D spectrogram; the values of compensation voltage in the position of PIP increased with the increase of humidity. T he results provided data support for optimizing the environmental humidity during the determination using FAIMS

Abstract: