Most Down Articles

    Published in last 1 year | In last 2 years| In last 3 years| All| Most Downloaded in Recent Month | Most Downloaded in Recent Year|

    Please wait a minute...
    For Selected: Toggle Thumbnails
    Barbalat Lemma and its application in analysis of system stability
    MIN Ying-ying,LIU Yun-gang
    Abstract551)      PDF(pc) (269KB)(7665)       Save
    A set of primary formulations of Barbalat Lemma and its simple alternatives are summarized. The relationships among those formulations and their applicable scopes are investigated. The applications of Barbalat Lemma in analyzing asymptotically convergence of the system, adaptive control design and Lp stability are discussed via three examples.
    Related Articles | Metrics
    Mechanical design and gait planning of a hydraulically actuated quadruped bionic robot
    LI Yi-bin1, LI Bin1,2, RONG Xue-wen1, MENG Jian1
    Abstract379)      PDF(pc) (1531KB)(5921)       Save

    A hydraulically actuated quadruped bionic robot has been developed by Center for Robotics at Shandong University is described in this paper. The objective is to design a highly dynamic and high load quadruped robot that enables the adaptation to complex terrain. Based on mule/horse creature bionics, the leg configuration with passive structure and hydraulic actuation are met the needs of stability control and high load capacity. And the stability dynamic trotting gait of the quadruped robot is planned based on the forward kinematics and inverse kinematics. Experiments of the developed quadruped bionic robot platform show the rationality of mechanical design and the effectiveness of gait planning.

    Related Articles | Metrics
    Wood defects recognition based on the convolutional neural network
    XU Shan-shan, LIU Ying-an*, XU Sheng
    Abstract472)      PDF(pc) (2379KB)(4926)       Save

    To improve the efficiency of wood defects identification, a method based on the convolutional neural network was proposed. A convolutional neural network was presented to recognize the wood defect, and the numbers of training samples were determined by an incremental learning method; the corresponding network structure was designed, and the time consumption could be reduced. Experimental results showed that the preprocessing of a complex image was not needed, and the multiclass defects could be recognized with high accuracy, small complexity and good robustness, while the inherent shortcomings of the traditional algorithm were overcame.

    Related Articles | Metrics
    DSP based implementation of the space vector pulse width modulation
    KONG Wei-tao,ZHANG Qing-fan,ZHANG Cheng-hui
    Abstract154)      PDF(pc) (1039KB)(4884)       Save
    According to the fundamental principle of motors, the space vector theory was analyzed, and a simple space vector pulse width modulation (SVPWM) method was proposed. The reference voltage space vector was directly synthesized based on the switching state vectors in α-β vectors complex space, and the duration time of each switching state vectors was calculated. The computational efficiency of this algorithm is moderate, and the realtime performance is elegant. Thus the current distortions and torque ripples are lower in the output of the inverter. This given method has been applied to a control system of the threephase induction motor based on TMS320LF2407A. The test results showed that this control system with SVPWM has such merits as simple realization, improved running performance and an enhanced d.c. voltage utilization ratio.
    Related Articles | Metrics
    Dynamic obstacle avoidance Algorithm for the mobile robot
    LI Cai-hong,LI Yi-bin,FAN Chen
    Abstract467)      PDF(pc) (298KB)(4618)       Save
    A new hybrid algorithm of dynamic obstacle avoidance was introduced for local path planning under an uncertainty environment. It is a combination of the rolling planning and RBF neural network (RBFNN) forecast. The moving trajectory of the dynamic obstacle was illustrated by using a camera lens, and a heart-shaped sequence was acquired from the samples. The RBFNN prediction model was built based on these data. A dynamic rolling window within the scope of the scanning ultrasonic sensor was established according to current location when the mobile robot was in real-time planning. Forecast computation was started when a rolling window into the dynamic obstacle was detected. The next moving location of the obstacle was predicted based on the three adjacent values of time sequence. Thus the dynamic obstacle avoidance issue converts into instantaneous static once and the real-time planning was reallized. This method can improve dynamic obstacle avoidance and the safety of real-time planning. Simulation results show that the method is feasible and efficient.
    Related Articles | Metrics
    A method on building Chinese sentiment lexicon for text sentiment analysis
    ZHOU Yong-mei1, YANG Jia-neng2, YANG Ai-min1
    Abstract1186)      PDF(pc) (1259KB)(4552)       Save

    A method on building Chinese sentiment lexicon based on HowNet and SentiWordNet was proposed,in which sentiment intensity of the word was automatically calculated by decomposing it into multiple semantic units and a lexicon proofreading technique was used to optimize the value of sentiment intensity of the word. The building lexicon was applied to the task of sentiment analysis, in which the support vector machine was used to build the sentiment classifier. The experiment results showed that the built sentiment lexicon was more effective than the general polar sentiment lexicon,and provided an effective dictionary resource for the research of sentiment analysis.

    Related Articles | Metrics
    Research on IDS-faced general-purpose application-level protocol identification technology
    LIU Yuan-xun,XU Qiu-liang,YUN Xiao-chun
    Abstract463)      PDF(pc) (295KB)(4515)       Save
    With the ceaseless developing of the computer network environment and strengthening of network application, it becomes more and more difficult to analyze protocol and detect intrusion. In addition, report mistakes and pretermission happen frequently. The reliability of protocol analysis and intrusion detection can be improved greatly by identifying applicationlevel protocol and analyzing different protocols with different methods before protocol analysis and intrusion detection. A generalpurpose applicationlevel protocol identification technology is proposed and applied to an intrusion detection system, whose usability and reliability are proved by the experimentation.
    Related Articles | Metrics
    The development status of hot stamping and quenching of ultra high-strength steel
    LI Hui-ping, ZHAO Guo-qun, ZHANG Lei, HE Lian-fang
    Abstract120)      PDF(pc) (652KB)(4366)       Save

    Hot stamping of ultra high-strength steels (UHSS) is a new and complex forming technology integrating metal hot forming and the quenching process. UHSS has many advantages due to hot stamping, such as low flow stress, good plasticity, and high forming limit. The shape accuracy and strength of parts formed using hot stamping are very high. So the strength of hot stamping parts satisfies the requirement of auto security, although the thickness of parts is reduced. In this paper, the types of UHSS, the key technologies of hot stamping were introduced. The main content and development status of hot stamping experimental study and numerical simulation study were summarized and reviewed, and the problems at present were also commented. Also, the content for further research and study direction were prospected.

    Related Articles | Metrics
    Analysis of perpetual pavement strain distribution and fatigue damage
    YANG Yongshun1, WANG Lin2, GAO Xuechi1, JIA Haiqing3, WEI Jincheng2
    Abstract150)      PDF(pc) (2517KB)(4198)       Save

    Because the HMA fatigue limit is the theoretical base of perpetual pavement design, to grasp the strain distribution and strain action of HMA layer is key for perpetual pavement design. The traffic, axle load spectrum and pavement temperature of a perpetual pavement test road was analyzed by field measured data. The characteristic of asphalt layers dynamic strain pulse was analyzed and regression models were constructed based on controlled axle load truck tests, which could correlate the maximum longitudinal tensile strain to pavement temperature and axle load. Based on the axle load spectrum, temperature distribution and strain regression models, the pavement strain distribution in a year was calculated. According to the fatigue limit theory, the cumulative damage method was used to calculate the pavement fatigue life, and the fatigue life of perpetual pavement test road was analyzed.

    Related Articles | Metrics
    A fast SVM-based feature selection method
    DAI Ping, LI Ning*
    Abstract124)      PDF(pc) (354KB)(4172)       Save

    Aiming at the large computation and slow convergence speed of the traditional feature selection methods, a fast SVMbased feature selection method is proposed to overcome.Support vecor machine is employed as the classifier and particle swarm optimization method is employed as searching strategy.The proposed method reduces the iterations of training classifiers by taking advantage of the characteristics of linear and polynomial kernel functions so that it reduces the complexity of calculation. Experimental results show that the method accelerates feature selection in the case of no loss of classification performances.

    Related Articles | Metrics
    Human motion detection based on multi-feature fusion
    TIAN Guo-Hui, JI Yan-Jing, HUANG Bin
    Abstract222)      PDF(pc) (792KB)(4124)       Save

    A novel method was proposed to detect human daily activities and  fall behavior based on features fusion. Actions were described by a set of postures and features of postures change velocity and position changes, then simple postures events and features were combined to express complex human activities events.This method can robustly detect human actions, thus can be used for intelligence interactive and service robot autonomous server system.

    Related Articles | Metrics
    A review of the research and development of biomass gasification technology in China
    DONG Yu-ping,DENG Bo,JING Yuan-zhuo,QIANG Ning,SHEN Shu-yun
    Abstract699)      PDF(pc) (370KB)(4079)       Save
    Biomass energy is an important renewable energy .The use of biomass gasification technology can achieve zero CO2 emissions, saving conventional energy, meeting the requirements of sustainable development. The paper introduces process characteristics of biomass gasification and related equipments; expounds on the direction of research in the fields of biomass gasification, analysis the status of the biomass gasification technology's commercialization in China. Finally, the proposed reference is put forward.
    Related Articles | Metrics
    A fast affinity propagation clustering algorithm
    LIU Xiao-yong1,2,3, FU Hui2
    Abstract158)      PDF(pc) (960KB)(3987)       Save

    An important parameter of affinity propagation algorithm (AP), damping factor, affects the speed of AP. Because the value of damping factor is fixed in traditional AP algorithm, the convergence performance of AP algorithm is sensitive to the parameter’s choosing. A novel and fast AP algorithm, F-AP, was proposed. The new algorithm used the constriction factor to regulate damping factor dynamically. Three datasets and iris dataset were used to compare AP and F-AP. The numerical results showed that F-AP could accelerate the convergence process effectively.

    Related Articles | Metrics
    Research and development of intelligent space technology for a home service robot

    TIAN Guo-hui,LI Xiao-lei,ZHAO Shou-peng,LU Fei

    Abstract535)      PDF(pc) (366KB)(3894)       Save
    Intelligent space technology is a new research field. The intelligent space technology for a home service robot can perceive men and objects in space more completely and realize the relevant service task by setting sensors and actuators distributely in the specified positions of space. Simultaneously the service robot can obtain more comprehensive environmental information through the intelligent space, instead of being equipped with too many sensors itself, and the service robot can work more rapidly, accurately and stably with light packs in an unknown or semiunknown dynamic environment. First, the research situation of intelligent space technology was summarized, then the structure, characteristics and key research topic for the intelligent space for a home service robot were introduced, and the relevant technologies were also proposed. Finally, the new research direction of the intelligent space for a home service robot was discussed.
    Related Articles | Metrics
    Abstract309)      PDF(pc) (398KB)(3838)       Save

    Abstract: The goal of super resolution is to get high resolution images from at least one low resolution image. However,

    with the traditional (interpolationbased and reconstructionbased) methods it is difficult to make further important

    progress. The emerging learningbased superresolution methods breathe new life into the research. By reviewing these three

    kinds of methods, the history of super resolution was surveyed, the limitation of existing methods was discussed and the

    roadmap was analyzed for future developments.

    Related Articles | Metrics
    A review of the research and development of biomass energy technology
    DONG Yu-ping,WANG Li-peng,DENG Bo,LU Ping,SHEN Shu-yun
    Abstract260)      PDF(pc) (365KB)(3785)       Save
    Biomass energy is a type of important renewable energy. It is very important for the solution of the energy crisis and the protection of environment. The work introduces the status quo of the research and application of biomass energy at home and abroad, expounds on the main aspects of the research and the existing problems, analyzes the obstacles in the research and exploration of the biomass in China, and some corresponding measures are proposed.
    Related Articles | Metrics
    Robot autonomous localization based on a single artificial landmark
    ZHAO Shou-peng,TIAN Guo-hui,LI Xiao-lei
    Abstract223)      PDF(pc) (293KB)(3704)       Save
    Based on four essential rules of how to design a landmark, a new style of landmark is proposed, that is a planar self-similar landmark based on QR code technology. The method to detect and recognize the landmark is studied. Based on camera calibration methods, the camera parameters are computed by finding the calibration points in the landmark, which can make the robot realize autonomous localization by only a single landmark. Experimental results show that the robot could locate and navigate fast and accurately using this style of landmark in a complex environment.
    Related Articles | Metrics
    Survey of outlier detection technologies
    CHEN Bin, CHEN Song-Can, PAN Zhi-Song, LI Bin
    Abstract450)      PDF(pc) (1025KB)(3475)       Save

    Outlier detection aims to detect those data that significantly deviate from the expected behavior, and thus is widely applied in many fields, such as, machine fault detection, intrusion detection, fraud detection and data preprocessing. Hence, there exist many generic and special algorithms for outlier detection under the unsupervised and supervised learning framework. But up to now, there still has been no clear classification in this aspect. To provide a structural view, the review of the state-of-the-art statistics-based methods for outlier detection was focusedon, and a simple classification was given in this aspect. Moreover,the equivalence between some outlier detectors in depth is particularly discussed.

    Related Articles | Metrics
    Abstract85)      PDF(pc) (554KB)(3433)       Save
    Related Articles | Metrics
    Traffic information forecast algorithm based on the one-dimension cloud model
    YU Shao-wei,CAO Kai,ZHAO Mo
    Abstract545)      PDF(pc) (270KB)(3411)       Save
    The cloud model is a mathematical representation of fuzziness and randomness in linguistic concepts. The qualitative concept is represented by expected value, entropy and hyper entropy in this model.and It can realize the transformation between the qualitative and the quantitative when the fuzziness and randomness of a linguistic concept are integrated together. Therefore, it has stronger robustness to the higher uncertain traffic information. A kind of traffic information forecast mechanism based on cloud model is proposed. This mechanism can overcome the fuzziness and the random in traffic information effectively, and can offer a novel approach for traffic information forecasting. It also shows that the predicted results are basically accord with the real ones.
    Related Articles | Metrics
    Adaptive spectral clustering algorithm
    BO De-Yun, ZHANG Dao-Jiang
    Abstract263)      PDF(pc) (1011KB)(3374)       Save

    Spectral clustering has been used to identify clusters that are non-linearly separable in input space, and usually outperforms traditional clustering algorithms. However, the performances of spectral clustering are severely dependent on values of the scaling parameter. In this paper, an adaptive spectral clustering (ASC) algorithm was proposed based on traditional spectral clustering, which can choose the scaling parameter automatically by using techniques similar to kernel selection. The new algorithm was compared to existing parameter selection based spectral clustering algorithms on both synthetic and UCI data sets, and the experimental results validate the effectiveness of the proposed algorithm.

    Related Articles | Metrics
    Study of pitch shifting technology and the sound quality evaluating
    ZHANG Xiao-rui, TIAN Lan*
    Abstract180)      PDF(pc) (1526KB)(3373)       Save

    Three pitch shifting methods were studied and implemented,which were the improved synchronized overlap-add fixed synthesis(SOLA-FS) method, interpolation-on-frequency method, and the phase vocoder method. The improved SOLA-FS method was especially discussed in detail. Principles and characteristics were analyzed and simulation results were provided. In the experiments, it was found that the three methods could all achieve the effects of pitch-shifting, that is the pitch of the audio would be changed but the duration would be kept unchanged. But the auditory perceptions of the three methods were different. So, a sound quality evaluation test was carried out. The results of auditory evaluation tests showed that, under the given pitch-shifting ratio, whether for up or down, the sound quality processed by the improved SOLA-FS method is the best.

    Related Articles | Metrics
    Application of PSPICE CAD in designing a highvoltage high power switching power supply
    CHEN Ya-fei,LEI Hong,CHEN Yin-xing,FENG Jin-jun
    Abstract432)      PDF(pc) (1045KB)(3345)       Save
    A highvoltage highpower switchingpowersupply employed for MicrowavePowerModule was designed and manufactured with the assistance of the PSPICE simulation. The modeling and simulating processes of design and optimization of the switchingpowersupply circuit for PSPICE is proposed. The factors that affect the accuracy and computer consuming time in switchingpowersupply PSPICE simulation are discussed. The specific methods to solve PSPICE convergence problem are concluded. Finally, the simulating output results are given with fairly good agreement when compared with the experimental results.
    Related Articles | Metrics
    Analysis of transmission spectra of 6H—SiC crystal in the visible
    and near infrared range
    YU Yuanxun1, LIAN Jie2*, GUAN Wenli1, WANG Gongtang3, LI Juan4, XU Xiangang4
    Abstract188)      PDF(pc) (819KB)(3308)       Save

    Transmission spectra of instinct and N doped 6H—SiC crystals in the visible and near infrared range were measured by using the UVVis spectrophotometer and infrared spectrum measurement system. Optical transmission spectra demonstrated that the 6H—SiC single crystal was transparent for visible and near infrared radiation. Compared with the 6H—SiC single crystal, it was also observed that  nitrogen doping could make the bandgap narrower, and the optical absorption of N doped SiC appeared in the visible and near infrared range. Also,  dispersion curves were obtained by the transmission spectra, and the refractive indices were analyzed using the dispersion equation in the transparent region. In addition,  spectra analysis showed that the absorption at 625nm corresponded to the free electron transition from the low energy level of band tailing induced by N doping to the high energy level in the conduction band, and the absorption of free carrier resulted in low transmission around 2500nm for nitrogen doped crystal.

    Related Articles | Metrics
    Head detection algorithm based on mathematical morphology and HSI color space
    ZHAO Jun-wei1, HOU Qing-tao2, LI Jin-ping3, PENG Bo4
    Abstract223)      PDF(pc) (1630KB)(3095)       Save
      An approach of head detection was proposed to meet the requirement of the reducing people counting time in real-time video monitoring. A hair color model was established in HSI color space, and the image binarization was processed by considering the range of hair gray-level. The mathematical morphology dilation and erosion preprocessing algorithm was presented. Then, dilation and erosion algorithm was used again after image edge detection in order to get several candidate areas. At last, the geometrical characteristics of contour were used to judge whether these areas head or not. Experimental results showed that this algorithm was fast and effective, which could meet the real-time needs of intelligent video surveillance.
    Related Articles | Metrics
    Preparation of biomimetic nanograde hydroxyapatite in simulated body fluid
    ZHANG Ai-juan
    Abstract465)      PDF(pc) (875KB)(3061)       Save
    Nanograde hydroxyapatite(HAP)powders were synthesized through the reaction between Ca(NO3)2 and Na3PO4 in simulated body fluid(SBF). The mass phase composition, chemical composition, micromorphology and the bioactivity of the obtained HAP were studied by X-ray diffraction, Fourier transform infrared spectroscopy, scanning electron microscope, X-ray photoelectron spectroscopy, and atomic absorption spectrometry. The Ca/P molar ratio of HAP was obtained by a chemical method. The results indicate that the synthesized HAP is a pinlike shape close to natural bone with 60~70?nm length, 10~20nm width, poor crystal structure and contains Na+ and CO2-3. The synthesized HAP powders have a good ability for inducing Ca and P deposition.
    Related Articles | Metrics
    Pose estimation based on semi-supervised latent Dirichlet allocation
    WEI Wei, ZHANG Yanning
    Abstract96)      PDF(pc) (983KB)(3042)       Save

    Topics cannot be labeled in the unsupervised topic model, while the labeling work in supervised topic models is tedious and subjective. To solve these problems, a semisupervised topic model was proposed. First, the locationirrelevant local features were detected and described by the scaleinvariant feature transform (SIFT), based on which images were represented by a bag of visual words. Then partial labels were introduced to the topicword level distribution in the latent Dirichlet allocation (LDA) model to guide the classification of the unlabeled data, which resulted in a semisupervised LDA (SSLDA) model. The validation on head pose estimation showed the classification rate of the proposed method was 9.0%~24.7% higher than that of LDA. And the pose classification rate on partially occluded and misaligned face images was 8.8% and 21.5%~39.8% higher than multi-pose PCA method. With a small amount of labeled images, the proposed SSLDA model approaches the fully supervised LDA method. And it is applicable to other image classification problems.

    Related Articles | Metrics
    A method of DOA estimation by a special array structure
    CHEN Lei,WANG Jian-ying,LV Xue,WANG Feng
    Abstract1096)      PDF(pc) (253KB)(2981)       Save
    A new method of DOA estimation based on a special array geometry arrangement is presented. By adding a sensor at a specific location on uniform linear array, a group of special sensor pairs can be obtained, which can break the ULA steering vector's periodicity. The special array not only overcomes the weakness of the ambiguity of DOA estimation, but also achieves the higher resolution than ULA at the same hardware cost. A matrixblock spatial smoothed algorithm applied to the special array is proposed. The data covariance matrix is divided into some blocks according to its steering vector's periodicity, then the traditional spatial smoothed algorithm is applied to each block. After reorganizing the data covariance matrix is amended, the DOA estimation of coherent source can be achieved.
    Related Articles | Metrics
    A  new algorithm of cold-start in a collaborative filtering system
    LI Gai1,2,3, LI Lei2,3
    Abstract173)      PDF(pc) (2031KB)(2961)       Save

     In the collaborative filtering algorithms based on matrix decomposition, the new user and new item cold-start is a difficult problem. The problem of cold-start was solved by using the attribute-to-feature a mapping algorithm based on K-nearest-neighbor(KNN) to get the feature vectors of the new user and new item. The experimental evaluation using a real-world dataset showed the effectiveness of this method.

    Related Articles | Metrics
    Development of a new rock similar material
    ZHANG Ning, LI Shu-Cai, LI Ming-Tian, YANG Lei
    Abstract163)      PDF(pc) (1137KB)(2957)       Save

    Based on the introduction of research status of similar materials commonly used in recent decades years at home and abroad, and advantages and shortcomings of these similar materials were analyzed. The physical properties of modified rubber powder cement mortar, including apparent density, uniaxial compressive strength, elastic modulus and splitting tensile strength, were studied. Variation of the physical parameters with replacement rate of rubber powder and water cement ratio was summarized, and a new rock similar material was produced. The results showed that the new rock similar material can well satisfy the requirements of similar material, and it was an ideal rock similar material and worth to be popularized in other model tests.

    Related Articles | Metrics
    SUN Dianzhu, ZHU Changzhi, LI Yanrui
    Abstract200)      PDF(pc) (596KB)(2874)       Save

    An improved extraction algorithm of boundary characteristic points was proposed, which includes four steps: first, the spacial

    index structure of the scattered pointcloudwas constructed based on the R*tree; second, the local model reference data was

    obtained based on the structure and the datum plane was set up; third, the maximum distance between the point of reference data

    and datum plane was computed, and the distance between the target point and the datum plane was computed; fourth, the data

    boundary characteristic was identified by comparing the two distances. The accuracy and rapidity extraction of scatter data

    boundary characteristic was proved by application examples.

    Related Articles | Metrics
    Construction of a fuzzy membership function based on interval number
    YU Shao-wei
    Abstract86)      PDF(pc) (524KB)(2831)       Save

    An approach based on interval number for constructing a fuzzy membership function considering some problems existing in statistical methods was proposed.The conception of the youth using interval number was described, the interval number was transformed into a limited boundary fuzzy figure based on the knowledge of interval number and fuzzy number, and then a α-cut set of a normal fuzzy number was obtained by weighted computing  extending it into a fuzzy membership function. Experimental results showed that this approach was simple and accurate with wide applicability and the fuzzy membership function was more continuous, objective and suitable to people's subjective consciousness.

    Related Articles | Metrics
    The structure and performances of the electroless Ni-P alloy coating on aluminum alloy
    SUN Cong-zheng,GUAN Cong-sheng,QIN Jing-yu,CHENG Chuan
    Abstract376)      PDF(pc) (262KB)(2827)       Save
    Ni-P electroless coating on aluminum alloy was plated by a double solution system of alkaline and acidic solutions, which was designed for proton exchange membrane fuel cell(PEMFC) bipolar plates. The content of the nickel and phosphorus in the coating was tested by the electron probe method, the influence of temperature and time of heat treatment on the crystallization structure by the X diffraction method, and corrosion resistance by the volt-ampere curve method. The experimental results show that heat treatment can obviously affect the crystallization structure and the corrosion resistance of the coating. The Ni-P coating with a P content of 12.1 %(wt) does not obviously crystallize after being treated with 200~250℃, while its corrosion resistance apparently improves. It can be used as PEMFCs' bipolar plates.
    Related Articles | Metrics
    Method of feature generation and selection for network traffic classification
    YANG Ai-min1, ZHOU Yong-mei1, DENG He2, ZHOU Jian-feng3
    Abstract97)      PDF(pc) (1195KB)(2816)       Save

    In the System of Network Traffic Classification based on machine learning method, feature generation and feature selection directly affects the speed and accuracy of classification. To solve this problem, in feature generation aspect, we analyze the packet’s attributes (size, count, time, flag) and flow’s attributes (time) from the information of Packet-Level and FlowLevel, and 37 statistical features are generated. In feature selection aspect, we proposes a method of feature selection integrating Filter model and Wrapper model, to decrease the dimension of features. Experiments show the proposed methods improve the accuracy of classification.

    Related Articles | Metrics
    Web spam detection based on SMOTE and random forests
    FANG Xiao-nan1,2, ZHANG Hua-xiang1,2*, GAO Shuang1,2
    Abstract192)      PDF(pc) (1009KB)(2798)       Save

    Web spam refers to the actions intended to mislead search engines into ranking some pages higher than they deserved, which could significantly deteriorate the quality of searching results. Considering the serious imbalance of the Web spam dataset, it was proposed to use oversampling method SMOTE to balance the dataset, then to train the classifiers with random forests algorithm. The results showed that the SMOTE+RF method was more prominent by means of experimental comparison with the conventional single classifiers and the ensemble learning classifiers. The important parameters of this method were optimized based on experimental results, and the reasons for the improvement of the AUC value after using SMOTE were also analyzed. Experimental results on WEBSPAM UK2007 dataset showed that this method could markedly improve the performance of the classifiers, of which the AUC value could exceed the best result of Web Spam Challenge 2008.

    Related Articles | Metrics
    Synthesis of bone-like hydroxyapatite in simulated body fluid
    ZHANG Ai-juan
    Abstract80)      PDF(pc) (973KB)(2793)       Save

    Bone-like nanograde hydroxyapatite(HAP)spicula were synthesized in simulated body fluid(SBF)by increasing the concentration of Ca and  P in the  SBF by a biomimetic method. The mass phase composition, chemical composition, micromorphology, thermal stability,  bioactivity and  high temperature  during stability at different sintering temperature 500~1300℃ of the obtained HAP were studied by X-ray diffraction, Fourier transform infrared spectroscopy, scanning electron microscope, differential thermal analysis and  atomic absorption spectrometry. The results indicated that the synthesized HAP had a  pin-like shape close to natural bone with a 70~80nm length and 10~20nm width. Though the degree of crystallization of HAP is better and better with  temperature increasing, the HAP could be decomposed when the sintering temperature is above 1000℃.The synthesized HAP powders have  good ability for inducing the deposition of Ca and P.

    Related Articles | Metrics
    Semantic similarity computation of Baidu encyclopedia entries based on SimRank
    YIN Kun, YIN Hongfeng*, YANG Yan, JIA Zhen
    JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE)    DOI: 10.6040/j.issn.1672-3961.2.2013.282
    The fuel quantity correction and calibration of the high pressure common-rail diesel engine
    WANG Hong-rong,ZHANG You-tong,WANG Jun,DING Xiao-liang
    Abstract384)      PDF(pc) (410KB)(2776)       Save
    The correction methods and matching strategy of fuel quantity were studied with the high pressure common-rail diesel engine. The influences of processing precision difference of the individual injector,  diversity of the injector driven current and the pressure fluctuation affecting the coherence of the injection quantity were analyzed. The injection fuel quantity was corrected by feed-back control of the driven current and compensation of the energizing pulse. By analyzing the  engine operation point, the matching strategy of fuel quantity on common-rail diesel engine was designed. The fuel quantity MAP was calibrated with an asymmetric step according to the frequency and importance of the engine operation. Finally, experiments were done on a test bench to verify the correction strategies. Experimental results show that the control and compensation strategies could improve the coherence of the injection fuel quantity.
    Related Articles | Metrics
    Preprocessing algorithms for Chinese handwriting recognition
    LI Ying,SONG Gang
    Abstract538)      PDF(pc) (265KB)(2759)       Save
    A series of complete and systematic algorithms on preprocessing of handwriting image is proposed for freeformat handwritten Chinese characters . The connected character removing is carried out by combining statistical characteristic of characters width with the gauss distribution. According to the location of valley point on projecting images, a threshold value is selected as best division point resulting to the easy realization of size normalizing, character mosaicand line slope correction. The proposed methods make the whole preprocessing subsystem more stable and perfect, which lay a dependable foundation for subsequent handwriting recognition.
    Related Articles | Metrics
    A new feature selection method for text categorization
    WANG Fa-bo, XU Xin-shun
    Abstract90)      PDF(pc) (342KB)(2756)       Save

    How to reduce feature dimension while maintaining categorization accuracy is a key issue of text categorization.  A new method based on information theory was proposed to solve this problem. This approach aims to eliminate sparsely distributed features and find features  useful for categorization. Working with these feature reduction methods, it could  further reduce the feature dimension. The performance of  this  proposed method was tested on benchmark text classification problems. The results showed that it could not only reduce the feature dimension to hundreds but also improve the performance.

    Related Articles | Metrics
    Kinetics of the adsorption of Reactive Brilliant Red K-2BP onto    modified wheat residue
    ZHONG Qian-qian, YUE Qin-yan*, LI Qian, LI Ying, XU Xing, GAO Bao-yu
    Abstract155)      PDF(pc) (1625KB)(2748)       Save

    For removing the dye from wastewater, the wheat residue was modified into an anion exchanger. The adsorption kinetics of Reactive Brilliant Red K-2BP (RBR K-2BP) from aqueous solutions onto modified wheat residue(MWR)were  studied in a batch reactor. Adsorption experiments were carried out as a function of pH and adsorbent dosage. In addition, the kinetic data were respectively fitted by pseudo-first-order equation, pseudo-second-order equation, Elovich equation and the intraparticle diffusion model. It was found that the pseudo-second-order equation could  provide  the best correlation to the data. The intra-particle diffusion was the main rate-controlling step for the adsorption of RBR K-2BP onto MWR. The adsorption activation energy of dye was 17.21kJ/mol, which indicated that the adsorption type  whole process was mainly of the physical adsorption.

    Related Articles | Metrics
    The design method and application of the non-inductive Chua′s circuit
    ZHANG Xin-guo1, XU Chong-fang1*, WANG Jin-shuang1, YAN Ji-cong1, HAN Ting-wu1,2
    Abstract100)      PDF(pc) (1309KB)(2739)       Save

    The classical Chua′s circuit design is difficult to accurately improve  because the presence of the inductor can cause  a big error. To solve this problem, a non-inductor chaotic circuit with the same output as the classical Chua′s circuit was presented. The corresponding static circuit structure(limiting-amplitude circuit)was found by deep analysis of the static characteristic curve of Chua’s diode. Combined with CNN technology, Chua′s circuit consisting of  a pure operational amplifier  was designed and two kinds of static nonlinear circuits used in Chua′s circuits were also given. Simulation results and physical experimental results showed that the design fully reaches anticipative objectives. The comparison results showed that  the Chua′s circuit with limiting-amplitude nonlinear circuits was the optimal design.

    Related Articles | Metrics
    A variable structure cascade double loop control to the startup process of the heavy truck
    LI Hong-bin,ZHANG Cheng-rui
    Abstract1063)            Save
    In order to enhance the smoothly character of the startup process of the heavy truck, a startup control algorithm is proposed based on the analysis of the acceleration fluctuation before and after the lock point, which adopts variable structure cascade double loop control to keep the rotating speed of the engine constant. Considering the nonlinear characteristics of the clutch operating system, the neuron adaptive PID control algorithm is adopted in the acceleration loop to solve the problem of nonlinear control. The experiment results show that acceleration fluctuation before and after the lock point can be restrained by using this algorithm, then the performance of the launch process is improved.
    Related Articles | Metrics
    Simulation system of proton exchange membrane fuel cells based on MATLAB
    WANG Zhen,HAN Ji-tian,WANG Ji-hao,ZHAO Wei
    Abstract639)      PDF(pc) (263KB)(2699)       Save
    A steady state mathematical model of proton exchange membrane fuel cell (PEMFC) is presented using empirical equations. Simulations and analysis are performed in MATLAB. The comparison between the modeling results and the experimental data shows good agreement. A simulation software is presented for simulating the characteristic parameters of PEMFC based on the MATLAB and Visual Basic software packages. The software is developed using the empirical models for PEMFC, SIMULINK, and Visual Basic. The simulation results show that the developed simulation software will be useful for the effective design, parameter analysis, and PEMFC demonstration.
    Related Articles | Metrics
    Influence of interfacial strength on fracture process and mechanical properties of fiber composites
    SUN Lili, JIA Yuxi, SUN Sheng, MA Fengde
    Abstract163)      PDF(pc) (393KB)(2685)       Save

    The interface is one of the key components of composite materials, which plays an important role in the fracture process and mechanical properties. In the current work, glass fiber surfaces were treated via the selfassembly monolayer method. The fracture process and mechanical properties of composites with different interfacial strength were investigated by a homemade tensile tester with an additional polarizing microscope. The results showed that in the case of a strong interface between the reinforcing fiber and matrix, the composites fracture in brittle mode, whereas in the case of a weak interface, the composites fracture in ductile mode. Furthermore, the reinforcing efficiency of the composite is related to the interfacial strength.

    Related Articles | Metrics
    Traffic flow prediction based on the wavelet neural network with genetic algorithm
    LI Jing-yu,LI Qi-qiang,HOU Hai-yan,YANG Li-cai
    Abstract666)      PDF(pc) (276KB)(2678)       Save
    For the highcomplexity, timevariation and probability of urban traffic flow, its realtime and exact prediction is critical to the research of intelligent traffic system, especially for the advanced traffic management system and advanced traveler information system. Based on the character of the traffic flow prediction, a GAWNN model is given based on the wavelet neural network with genetic algorithm. The genetic algorithm of natural evolving law for the gradient descendent algorithm in Wavelet Neural Network is partly substituted to preoptimize the connection weight and the extension scale of the wavelet neural network, and later optimize the parameters along single gradient vector. This method overcomes some drawback when there exists single gradient descendent algorithm, such as local minimum and oscillation. A shorttime traffic flow prediction simulation using the GAWNN prediction model demonstrates the validity of the model.
    Related Articles | Metrics
    Distributed affinity propagation clustering algorithm based on GraphLab
    CHEN Wen-qiang1, LIN Chen1,2, CHEN Ke3, CHEN Jin-xiu1, ZOU Quan1,2*
    Abstract281)      PDF(pc) (1394KB)(2675)       Save

    A distributed affinity propagation algorithm based on GraphLab was proposed, which was named GStrAP (Graphlab based stream affinity propagation). In GraphLab′s DAG abstraction, the parallel computation was represented as a directed acyclic graph with data flowing along edges between vertices, and the “Gather-Apply-Scatter” paradigm was applied to complete data synchronization and algorithm′s iteration. The experimental results on 3D Clusters, Aggregation, Flame and Pathbased datasets with different scale and the clustering performance were compared with Kmeans, which demonstrated that the proposed GStrAP could achieve high performance on both scalability and accuracy.

    Related Articles | Metrics
    Dynamic-EIQ-ABC analysis application in distribution center programming
    SUN Lei,WU Yao-hua,ZHANG Guannv
    Abstract331)      PDF(pc) (266KB)(2657)       Save
    Based on the traditional EIQ analysis, dynamicEIQABC analysis application in the area of distribution center's programming is proposed. In addition, use of the DynamicEIQABC analysis method is also put forward in details. Then, the influences of different curve trends on distribution center programming and the type choosing of picking up devices is analyzed, in order to provide an efficient diagnostic method for distribution center running. Simulation analysis and the real running results prove that this method can well satisfy the Stockin/Stockout requests of the distribution center with an economic character and necessary free space for future development.
    Related Articles | Metrics
    Traffic volume forecasting algorithm based on RBF neural network
    ZHU Wen-xing,LONG Yan-ping,JIA Lei
    Abstract294)      PDF(pc) (286KB)(2649)       Save
    Traditional training algorithms for radial basis function (RBF) neural networks usually start with a predetermined hidden layer structure, which is selected by using a priori knowledge and based on previous experience. The parameters of RBF networks are trained by using traditional clustering and the least squares method. These training algorithms are always based on the local search method and often suffer from being trapped at structural local minima. A new method for training RBF structural parameters by using a genetic algorithm is put forward, of which a new clustering method named the sorting MSS (most scattered set) clustering method and a traditional least square method are incorporated. This method can not only prevent the result of the network from being trapped at local minima but also highly improves the computational efficiency. It gives satisfactory results when this algorithm is applied to traffic flow forecasting.
    Related Articles | Metrics
    Research status and developing trend analysis of the water inrush mechanism for underground engineering construction
    LI Li-ping1, LU Wei1,2, LI Shu-cai1, ZHANG Qing-song1, XU Zhen-hao1, SHI Shao-shuai1
    Abstract97)      PDF(pc) (1906KB)(2647)       Save

    With the developing trend of long, large and deep construction characteristics for underground engineering, larger water inrush disaster becomes the bottleneck problem for the further development of underground engineering. Based on an extensive literature review, the status and developing trend of the water inrush mechanism and countermeasures were summarized in detail, such as catastrophe condition and evolution characteristics of water inrush, formation mechanism of water inrush passage, and infestation regularity and mutation mechanism of seepage flow, based on which, a new academic idea has been proposed that artificial disturbance of excavation and disturbance of karst water and its pressure leads to water inrush of relatively intact rock masses, and the polyphyletic dynamic information of seepage mechanics for geological defects is the key for the catastrophe evolution description of geological defects. Based on  example analysis and numerical simulations, two aspects must be further studied, which are the microcosmic effect mechanism of the water inrush passage formation and macroscopic instability criterion for the system of water bearing structure considering the effect of geological factors and engineering effort factor, and the geophysical method effectiveness for the water bearing structure diction can be improved by combining with the monitoring method of polyphyletic information for water inrush prediction, which will lay a theoretical basis for the groundwater inrush prediction in underground engineering.

    Related Articles | Metrics