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|

    In last 3 years
    Please wait a minute...
    For Selected: Toggle Thumbnails
    Review and prospect of the development of heat exchanger structure
    Wenjing DU,Junzhe ZHAO,Lixin ZHANG,Zhan WANG,Wanxiang JI
    Journal of Shandong University(Engineering Science)    2021, 51 (5): 76-83.   DOI: 10.6040/j.issn.1672-3961.0.2020.423
    Abstract2146)   HTML148)    PDF(pc) (4892KB)(1341)       Save

    The development of heat exchanger for more than 200 years was introduced. The generation and typical applications of heat exchangers with different structures were reviewed. Four heat exchangers with different structures including shell-and-tube heat exchanger, plate heat exchanger, microstructure heat exchanger, and printed circuit heat exchanger were described emphatically, and the development work of related geometric parameter optimization and structure improvement was carried out to realize heat transfer enhancement. The existing problems and limitations in the structure design of the heat exchanger were analyzed, and the specific suggestions and development trends for the structure improvement of the heat exchanger in the future were proposed.

    Table and Figures | Reference | Related Articles | Metrics
    Review and prospect of research on power system inertia with high penetration of renewable energy source
    Hengxu ZHANG,Zhimin GAO,Yongji CAO,Hao QIN,Dong YANG,Huan MA
    Journal of Shandong University(Engineering Science)    2022, 52 (5): 1-13.   DOI: 10.6040/j.issn.1672-3961.0.2022.237
    Abstract1264)   HTML75)    PDF(pc) (4575KB)(1132)       Save

    With the aim of carbon peak and carbon neutrality, the integration of high-proportion renewable energy sources (RESs) makes the low-inertia characteristic of the new power system more obvious. In order to ensure power system security and stability, and support the integration of RESs, the research on the power system inertia was reviewed and prospected. The essence of the power system inertia was introduced, and the correlation among the conventional inertia, virtual inertia and equivalent inertia was discussed, on which the physical significance of the virtual inertia was revealed. From the viewpoints of the sources of inertia, the research on the virtual inertia control of the wind machine, photovoltaic generation, and energy storage system was summarized. Additionally, the assessment methods of power system inertia were reviewed. The important issues to be focused in the research area of power system inertia were summarized, and the suggestions for further studies were provided.

    Table and Figures | Reference | Related Articles | Metrics
    Review on development of simultaneous localization and mapping technology
    Jianqing WU,Xiuguang SONG
    Journal of Shandong University(Engineering Science)    2021, 51 (5): 16-31.   DOI: 10.6040/j.issn.1672-3961.0.2021.168
    Abstract1876)   HTML254)    PDF(pc) (2986KB)(1083)       Save

    As a hot spot in the field of intelligent transportation, simultaneous localization and mapping (SLAM) technology is the key to autonomous path planning for self-driving vehicles. This review focused on four parts with introduction of sensors related to SLAM technology, localization, mapping, and multi-sensor integration. Each step of realization for SLAM technology was introduced from advantages and disadvantages, range of application, probability algorithm, types of map, and integration methods. Based on the investigation of relevant researches about multi-sensor integration, common problems of SLAM technology were analyzed, future development trend and practical engineering application of SLAM technology were prospected.

    Table and Figures | Reference | Related Articles | Metrics
    Research progress of ocean wave energy converters
    Yanjun LIU,Shuang WU,Dengshuai WANG,Ruohong WANG
    Journal of Shandong University(Engineering Science)    2021, 51 (5): 63-75.   DOI: 10.6040/j.issn.1672-3961.0.2021.234
    Abstract2321)   HTML130)    PDF(pc) (13090KB)(919)       Save

    In order to accurately keeping the developments of ocean wave energy utilization technology, the important research progress of wave energy converters was reviewed. The forms of wave energy converters had not yet converged. According to the classification principle of different energy capture methods, the working principle and the energy conversion system of three main types of devices, namely, oscillating water column, overtopping and oscillating bodies were introduced. The advantages and disadvantages of each type of converter were analyzed. Typical engineering devices were selected for detailed introduction. The engineering devices which had been completed sea trial in China were summarized. It has been found that the oscillation type was the most popular type of converter in China. The research progress related to the performance evaluation of wave energy converters were summarized, but there was no unified evaluation standard yet. The difficulties and the main breakthrough directions of the developments of wave energy converters were discussed from three aspects of high efficiency and stability, reliability and cost, and the construction of diversified integrated platforms.

    Table and Figures | Reference | Related Articles | Metrics
    Automatic detection research of arrhythmia based on CNN-LSTM hybrid model
    TAO Liang, LIU Baoning, LIANG Wei
    Journal of Shandong University(Engineering Science)    2021, 51 (3): 30-36.   DOI: 10.6040/j.issn.1672-3961.0.2020.445
    Abstract1415)      PDF(pc) (3912KB)(789)       Save
    A hybrid algorithm of convolutional neural network and long short-term memory network was proposed for automatic detection of arrhythmias. The model structure was composed of 5 convolutional layers, 5 pooled layers, 1 LSTM layer and 1 fully connected layer. By taking advantage of CNN's ability to automatically extract features and LSTM's ability to capture dependencies before and after time series, the simple preprocessed ECG signal data were directly input into the hybrid model. The whole model combined the two steps of feature extraction and classifier classification, so as to identify five different arrhythmias more efficiently and accurately. The accuracy, sensitivity and specificity of the test set were 99.48%, 99.47% and 99.86% respectively. The experimental results showed that the proposed method could efficiently and accurately identify different types of arrhythmias.
    Reference | Related Articles | Metrics
    Eye tracking in human-computer interaction control
    Hui HE,Junhao HUANG
    Journal of Shandong University(Engineering Science)    2021, 51 (2): 1-8.   DOI: 10.6040/j.issn.1672-3961.0.2020.346
    Abstract1654)   HTML80)    PDF(pc) (4362KB)(730)       Save

    To actualize the simple and low-cost eye-tracking based human-computer interaction, an exact interaction method based on the visual directions estimation and eye tracking with webcam videos was proposed. A simple and fast convolution neural network model was used to roughly estimate the user′s viewpoints on the screen. And then an accurate human-computer interaction method was proposed on the basis of the eye movements recognition and sight line tracking results. To verify the effectiveness of the method, the key operations of eye mouse and eye typing were developed. The test results show that the proposed method enabled users to achieve eye tracking and to actualize most precise human-computer interactions with only one common monocular camera, which was expected to completely replace the mouse and keyboard hardwares.

    Table and Figures | Reference | Related Articles | Metrics
    Comprehensive protection scheme for grounding fault in low resistance grounding system
    HUANG Fuquan, WANG Tinghuang, ZHANG Haitai, LIU Zijun, LI Guodong
    Journal of Shandong University(Engineering Science)    2021, 51 (3): 113-118.   DOI: 10.6040/j.issn.1672-3961.0.2020.159
    Abstract1511)      PDF(pc) (2199KB)(656)       Save
    Because the existing grounding protection in low resistance grounding system had imcomplete protection configuration, poor selectivity, low sensitivity and insufficient ability to detect high-imedance grounding fault, the characteristics of zero sequence current while single-phase grounding fault occured in low resistance grounding system was analyzied. With the help of longitudinal cooperation between the upper and lower protections, the multistage grounding protection based on zero sequence current and the high sensitivity grounding protection based on low setting and time delay were proposed, and also the grounding protection configuration scheme and the setting principles at all levels were discussed. Fault line selection method was proposed for the high-impedance grounding fault using the lateral comparasion of the amplitudes of zero sequence currents between the outlet of each feeder and the neutral line. The feasibility and reliability of the proposed comprehensive protection scheme for grounding fault were verified by the simulation in a typical small resistance grounding distribution network.
    Reference | Related Articles | Metrics
    Algorithm of adaptive slope adjustment of quadruped robot based on model predictive control and its application
    LIANG Qixing, LI Bin, LI Zhi, ZHANG Hui, RONG Xuewen, FAN Yong
    Journal of Shandong University(Engineering Science)    2021, 51 (3): 37-44.   DOI: 10.6040/j.issn.1672-3961.0.2020.320
    Abstract1350)      PDF(pc) (6868KB)(521)       Save
    In order to realize the adaptive and stable walking of the quadruped robot on slope terrain, the adaptive adjustment strategies of both feet position and trunk posture of quadruped robot on slope were proposed based on the model predictive control. The posture determination parameters of the robot in locomotion were measured by the inertial measurement unit(IMU). By means of the derived foot end trajectory algorithm, the coordinate mapping of its toe position was obtained in order to adjust the center of gravity of the robot on the slope. Then the adaptive adjustment of the trunk posture of the robot in the process of climbing could be achieved through the trunk posture adjustment algorithm by means of designed “virtual slope”. With the help of the physical platform of quadruped robot and the actual slope terrain environment built in the laboratory, the feasibility and validity of the proposed algorithm are verified. Experimental result showed that the proposed slope adaptive control method had improved the stability margin of the robot on the slope and optimized the foot end motion space, thus leading to the realization of adaptive adjustment in climbing slope for the quadruped robot.
    Reference | Related Articles | Metrics
    Photovoltaic power prediction method based on NWP irradiance inclination conversion
    Wenling JIANG,Yanqing ZHAO,Bo WANG,Shuanglei FENG,Yan PEI,Fei ZHANG
    Journal of Shandong University(Engineering Science)    2021, 51 (5): 114-121.   DOI: 10.6040/j.issn.1672-3961.0.2020.104
    Abstract1218)   HTML28)    PDF(pc) (2582KB)(434)       Save

    Short-term photovoltaic (PV) power predictions statistical methods generally used the horizontal irradiance in numerical weather prediction (NWP) directly, but not the inclined irradiance received by tilt PV panels, which led to the calculation results not accurate enough. This paper proposed a new method to convert horizontal irradiance to inclined one. The method distinguished scattering into molecular scattering and Mie scattering with different properties, and then converted them separately. Using the new method, the paper converted NWP horizontal irradiance to inclined one. Based on inclined one, PV power model and prediction were made. The result of an example showed that the new method, whose root mean square error was 10.25% and correlation coefficient was 0.914 0, was more accurate than the traditional method, which used NWP horizontal irradiance directly.

    Table and Figures | Reference | Related Articles | Metrics
    Forward modeling analysis of seismic wave field for TGS360pro advanced prediction of groundwater: taking the karst model as an example
    ZHANG Mingcai, JU Guanghong, XIONG Zhangqiang, ZHANG Dazhou
    Journal of Shandong University(Engineering Science)    2021, 51 (3): 68-75.   DOI: 10.6040/j.issn.1672-3961.0.2020.158
    Abstract1109)      PDF(pc) (8916KB)(429)       Save
    In order to avoid the occurrence of water gushing accident due to the inaccurate detection of the underground water storage state in front of the tunnel during the excavation of the tunnel, the tunnel geology survey TGS360pro System was used to accurately and efficiently detect the water-bearing body in front of the tunnel face by using the relation between the instantaneous amplitude and frequency of seismic wave signal and the rock structure plane, at the same time, based on the principle of TGS360pro tunnel Prediction System, the geological model with water-filled cave in front of the face was simulated by using spectral element method, through the stress curve, the numerical analysis results showed that the TGS360pro Tunnel Prediction System could effectively detect the water-rich condition in front of the tunnel, it had the prospect of popularization and application in tunnel excavation construction.
    Reference | Related Articles | Metrics
    Review and prospect of renewable energy ramp events
    Xueshan HAN,Xinyi WANG,Ming YANG,Yixiao YU
    Journal of Shandong University(Engineering Science)    2021, 51 (5): 53-62.   DOI: 10.6040/j.issn.1672-3961.0.2020.287
    Abstract1936)   HTML56)    PDF(pc) (2283KB)(423)       Save

    This article elaborated the wind and solar power ramp events from the three levels: the definition, prediction method and control strategy. The common definitions of wind and solar ramp were listed, and their advantages and disadvantages were analyzed in turn. It summarized the current mainstream prediction methods of ramp events, which were divided into direct method and indirect method according to whether the power prediction results were used, and the commonly used evaluation indicators were evaluated. It elaborated the current commonly used methods of control strategy, taking energy storage participation as a division, it was divided into finite control strategy without energy storage participation and infinite control strategy with energy storage participation. The current research problems and key research directions in the future were summarized and prospected.

    Table and Figures | Reference | Related Articles | Metrics
    Research progress on preparation methods of engineering nano material particle aerosol
    Hui HU, Ningye TANG, Siyu WANG, Haisen WANG
    Journal of Shandong University(Engineering Science)    2022, 52 (4): 1-11.   DOI: 10.6040/j.issn.1672-3961.0.2022.009
    Abstract753)   HTML38)    PDF(pc) (3186KB)(379)       Save

    With the development of nanotechnology and biochemistry, higher requirements were put forward for the particle size and manufacturing process of micro particles. Aerosol, as a common engineering nano material particle, was widely used in all walks of life. The research progress of common aerosol preparation methods, which starts from the aerosol preparation methods and systematically based on a large number of relevant literatures, the applicable scenarios of different aerosol preparation methods, the application requirements were summarized of different aerosol preparation methods were defined, the advantages and disadvantages of were compared and analyzed different aerosol preparation methods. On this basis, the optimization and improvement of various methods by domestic and foreign scholars in different periods were further ummarized, and the experimental and research results of domestic and foreign scholars were deeply analyzed, and the application of supercritical fluid technology in aerosol preparation was analyzed and prospected.

    Table and Figures | Reference | Related Articles | Metrics
    Real-time semantic segmentation of high-resolution remote sensing image based on multi-level feature cascade
    Chunhong CAO,Hongxuan DUAN,Ling CAO,Lele ZHANG,Kai HU,Fen XIAO
    Journal of Shandong University(Engineering Science)    2021, 51 (2): 19-25.   DOI: 10.6040/j.issn.1672-3961.0.2020.225
    Abstract1210)   HTML40)    PDF(pc) (4639KB)(378)       Save

    Aiming at the problems of long segmentation time and inaccurate segmentation of small targets in remote sensing image semantic segmentation, a fast semantic segmentation model of high-resolution remote sensing image based on multi-level feature cascade network (MFCNet) was proposed. The model was mainly composed of feature encoding, feature fusion and target refinement. Feature encoding extracted the input images feature of different resolutions and used different backbone networks. Due to the lower resolution of low-resolution images, heavy-weight backbone networks were used to obtain rich semantic information with fewer parameters. For medium and high-resolution images, lightweight backbone network was used to reduce the amount of parameters and obtain global information. While medium and low-resolution encoding used the way of weights and calculation sharing to further reduce model parameters and computational complexity. The feature fusion section fused features from different branches to obtain information at different scales. The target refinement used residual to correction the fused features and the features of the coded part to restore the spatial detail information of the image, making the segmentation more accurate. And the entire model worked efficiently in an end-to-end manner. The experimental verified the validity of the model in semantic segmentation of remote sensing images, and achieved a good balance between model complexity and accuracy.

    Table and Figures | Reference | Related Articles | Metrics
    Optimal calculation method of pavement shakedown limit based on genetic algorithm
    Xiuguang SONG,Yingchao ZHANG,Peizhi ZHUANG,He YANG,Haifeng ZHANG,Juan WANG
    Journal of Shandong University(Engineering Science)    2021, 51 (5): 1-7.   DOI: 10.6040/j.issn.1672-3961.0.2021.153
    Abstract1070)   HTML324)    PDF(pc) (4005KB)(362)       Save

    To solve the problem that road structures are prone to elasto-plastic deformation under the action of long-term reciprocating vehicle loads, based on the static shakedown theorem, the shakedown behavior of the semi-infinite space Mohr-Coulomb structure under the Hertz load was studied, and the genetic algorithm was introduced to construct an efficient calculation method for the lower limit of the shakedown limit of the road structure under the reciprocating vehicle load. The accuracy and efficiency of the new method was verified by comparison with the existing solution method and parameter analysis.

    Table and Figures | Reference | Related Articles | Metrics
    Review of risk conflict identification and early warning for interchange
    Jianqing WU,Qifeng WANG,Zhouyuan LI,Yuan TIAN
    Journal of Shandong University(Engineering Science)    2022, 52 (6): 1-13.   DOI: 10.6040/j.issn.1672-3961.0.2022.172
    Abstract761)   HTML28)    PDF(pc) (2835KB)(357)       Save

    This review focused on three parts with data collection method, related indexes of traffic conflict and conflict risk early warning method. Each step of the technology was introduced from range of application, working principle, advantages and disadvantages.Based on the real-time multi-sensor data fusion and roadside early warning method, the future development trend and practical application of risk conflict identification and early warning technology for urban interchanges were prospected.

    Table and Figures | Reference | Related Articles | Metrics
    Optimization of ventilation system of TBM tunnel construction and evaluation of dust suppression effect
    WANG Chunguo
    Journal of Shandong University(Engineering Science)    2021, 51 (3): 52-60.   DOI: 10.6040/j.issn.1672-3961.0.2020.468
    Abstract1062)      PDF(pc) (5863KB)(337)       Save
    Hard rock tunneling boring machine(TBM)produces a large amount of dust during construction, which is an important factor affecting the operating environment and the health of workers. To further optimize the construction ventilation and dust removal effect, combined with the actual working conditions of Qingdao Metro Line 1 double shield TBM, Ansys-Fluent software was used to carry out numerical analysis of the tunnel excavation process. The wind speed and dust mass concentration at each location of the tunnel were detected and compared with the numerical simulation results to verify the effectiveness of the model. In view of whether it is necessary to open the dust removal system in the process of TBM tunnel construction, as well as the location of dust removal tuyere and the selection of optimal suction flow, numerical simulation was carried out. When the dust removal system was turned off or the dust removal system was turned on but the suction flow was below 4 m3/s, and the dust diffuses to most areas of the TBM tunneling area. When the dust duct was 15 m away from the hand surface and the suction airflow was 12 m3/s, the dust removal effect reached the best, and the dust diffusion distance was reduced to 45 m, which could effectively remove dust. The research results could provide a scientific basis for the design and construction of tunnel ventilation and dust removal.
    Reference | Related Articles | Metrics
    Recognition learning based on multivariate functional principal component representation
    Yinfeng MENG,Qingfang LI
    Journal of Shandong University(Engineering Science)    2022, 52 (3): 1-8.   DOI: 10.6040/j.issn.1672-3961.0.2021.314
    Abstract642)   HTML8)    PDF(pc) (6668KB)(324)       Save

    Aiming at the problem of multi-dimensional information fusion in recognition learning, a recognition method based on multivariate functional principal component representation was proposed. The numerical calculation method of multivariate functional principal components was given. The joint covariance operator was used to calculate eigenvalues and eigenvectors, and the key distinguishing features were extracted. Based on these comprehensive features, the random forest method was used to recognize and learn multivariate functional data. The recognition performance of multivariate functional principal component representation method was compared with other representation methods on simulated data and real data. The experimental results showed that the accuracy was equal to 1 in the simulation dataset, English handwritten dataset and Chinese handwritten dataset, and 0.954 4 in the motion dataset. Compared with other methods, multivariate functional principal component analysis (MFPCA) had better recognition effect and improved the recognition accuracy effectively.

    Table and Figures | Reference | Related Articles | Metrics
    Small sample person re-identification combining Gabor features and convolution features
    FU Guixia, ZOU Guofeng, MAO Shuai, PAN Jinfeng, YIN Liju
    Journal of Shandong University(Engineering Science)    2021, 51 (3): 22-29.   DOI: 10.6040/j.issn.1672-3961.0.2020.232
    Abstract839)      PDF(pc) (6150KB)(322)       Save
    In the video surveillance, the limited available person images and unreliable data annotation led to the performance degradation of supervised person re-identification. To solve these problems, we proposed an unsupervised small sample person re-identification method that integrated Gabor features and convolution features. Gabor transform was used to extract multi-scale and multi-direction person texture and edge information, so as to realize the data augmentation of small sample person images in feature level. The redundant information was eliminated by feature encoding to improve the efficiency of feature similarity calculation. The convolutional auto-encoder network was adopted to extract the nonlinear deep convolution feature of pedestrian, which avoided the dependence of supervised learning algorithm on data annotation. The fusion of two heterogeneous features was applied to person similarity comparison, which implemented the feature augmentation of small samples and the improvement of person feature discrimination ability. Experiments were implemented based on Market-1501 and DukeMTMC-reID datasets, the rank-1 accuracy reached 74% and 67.1% respectively. The experimental results showed that the proposed network framework effectively improved the performance of small sample person re-identification.
    Reference | Related Articles | Metrics
    Location optimization of median BRT station at intersection
    GUO Rongrong, ZHANG Ruhua, MA Xinhui, GUO Senyao
    Journal of Shandong University(Engineering Science)    2021, 51 (3): 61-67.   DOI: 10.6040/j.issn.1672-3961.0.2020.345
    Abstract810)      PDF(pc) (1618KB)(320)       Save
    Based on the optimal comprehensive time efficiency of BRT vehicles and arriving passengers, a station selection optimization method was proposed. The time efficiency model was established and calculated by fuzzy analytic hierarchy process and genetic algorithm. Huanggang Road Station in Jinan was selected for case analysis. The results showed that the larger the weight of BRT operation time in a measurement unit was, the smaller the weight of arrival passenger travel time was, and the greater the buffer distance when the time efficiency was optimal. When the case station was set up at 25 meters downstream of the intersection, the efficiency of comprehensive time was the most optimal, which was 20% higher than the current situation.
    Reference | Related Articles | Metrics
    Pollution characteristics and intake risk assessment of short and medium-chain chlorinated paraffins in foods in Jinan
    Xinxin FANG,Shiwen ZHANG,Yuting ZHU,Wei JIANG,Zhaoyuan ZHANG,Nan ZHAO
    Journal of Shandong University(Engineering Science)    2021, 51 (3): 119-128.   DOI: 10.6040/j.issn.1672-3961.0.2021.001
    Abstract893)   HTML182)    PDF(pc) (3853KB)(314)       Save

    To explore the distribution status and intake risk of short-chain chlorinated paraffins (C10-13, SCCPs) and medium-chain chlorinated paraffins (C14-17, MCCPs) in food in Jinan city, 82 kinds of food were collected in Jinan in 2020 and analyzed by chlorine enhanced atmospheric pressure chemical ionization source- four-stage pole time-of-flight mass spectrometry. The results showed that the concentration of wet weight of SCCPs in food samples ranged from 5.3 to 2 483.2 ng/g, MCCPs ranged from 4.6 to 605.1 ng/g. The concentration of wet weight of SCCPs in Peanut oil was the highest, which was 2 115.5 ng/g. The concentration of wet weight of SCCPs in balsam pear was the lowest, which was 5.7 ng/g. The concentration of wet weight of MCCPs in soybean oil was the highest, which was 605.5 ng/g. The concentration of wet weight of MCCPs in Chinese cabbage was the lowest, which was 6.2 ng/g. The estimated daily intake (EDI) of SCCPs was the highest in staple foods, which was 2 619.2 ng/(kg ·d) and the lowest in aquatic foods, which was 17.7 ng/(kg ·d). The EDI of MCCPs was the highest in staple foods, which was 2 117.6 ng/(kg ·d), and the lowest in aquatic foods, which was 12.1 ng/(kg ·d). The hazard quotients (HQ) of SCCPs、MCCPs was 0.041 and 0.032, respectively. The purpose of this study was to provide reference data for the assessment of the risk of exposure to CPs among the population in Jinan.

    Table and Figures | Reference | Related Articles | Metrics
    Image-dependent fusion method for saliency maps
    Ye LIANG,Nan MA,Hongzhe LIU
    Journal of Shandong University(Engineering Science)    2021, 51 (4): 1-7.   DOI: 10.6040/j.issn.1672-3961.0.2020.266
    Abstract723)   HTML33)    PDF(pc) (4939KB)(303)       Save

    A saliency fusion method based on ridge regression was proposed to obtain better detection performance. The nearest neighbor set of the image to be detected was searched in the training set. The ridge regression method was used to estimate the fusion coefficients of different saliency maps. The saliency maps of different detection methods were fused. This method fully considered the differences of detection methods, and solved the problem of saliency map fusion in the absence of benchmark binary annotations. The AUC value of the proposed method was 0.911 on ECSSD dataset. The AUC value of the proposed method was 0.987 on HKU-IS dataset. The AUC value of the proposed method was 0.953 on DUT-OMRON dataset. The efficiency of the proposed method was verified by experimental results.

    Table and Figures | Reference | Related Articles | Metrics
    Modified calculation method of shaft friction for driven pile considering particle size effect
    Peizhi ZHUANG,Yingchao ZHANG,Xiuguang SONG,He YANG,Zhicheng GUO,Yan HU
    Journal of Shandong University(Engineering Science)    2021, 51 (5): 8-15.   DOI: 10.6040/j.issn.1672-3961.0.2021.161
    Abstract932)   HTML229)    PDF(pc) (7930KB)(301)       Save

    This study aimed to investigate the influence of particle size on the micropile by conducting theoretical analysis and model tests. The empirical relationship between the critical friction angle and the relative roughness at the pile-soil interface was established, and thus the critical friction angle could be determined quantitively in consideration of the particle size of sands. To emphasis the influence of particle size on the additional normal stress at the soil-pile interface, the shear band at the soil-pile interface was modelled as a hollow cylinder and then a new modified method was proposed based on the elastic cavity expansion theory. Only two new parameters, Poisson's ratio and the thickness of the shear band, were involved in the modified method, which was validated by comparison with model tests. It was found that the pile shaft friction was mainly determined by the pile roughness and the ratio of pile diameter to sand median size and the critical state angle at the pile-soil interface, while the additional normal stress mainly results from the pile roughness and the ratio of pile diameter to sand median size, respectively. The research could provide the valueable reference for the bearing capacity design of micropiles.

    Table and Figures | Reference | Related Articles | Metrics
    Analysis of the influence of the underside baffle and deflector of the fume hood on the flow field
    Ruiyi YAN,Zhen DONG,Sen LU,Yanhua LAI,Mingxin LÜ
    Journal of Shandong University(Engineering Science)    2021, 51 (5): 122-130.   DOI: 10.6040/j.issn.1672-3961.0.2021.177
    Abstract928)   HTML29)    PDF(pc) (19825KB)(301)       Save

    For the phenomenon of vortex and gas escape in the flow field of the extant fume hood, the computational fluid dynamics method was used for simulation. This research changed the shape parameters of the underside structure of the fume hood window, used circular arc baffles with different radii, elliptical arc baffles with different lengths, and added deflectors with different radii. Then this research studied the influence of structural changes on the flow field in the fume hood by analyzing the size and location of the vortex to obtain the direction of structural optimization. The results showed that when the height of the underside baffle was low and the length was short, there was obvious large vortex in the center position of the fume hood and near the wall. A baffle with a height greater than 50 mm and a length greater than 90 mm could significantly reduce the vortex at the center of the fume hood and at the junction of the baffle and the side wall. Adding a deflector could optimize the flow field and make the center vortex disappear, and the radius of the deflector should not be greater than 70 mm, otherwise obvious vortices would be excited on the outside and end of the baffle, which provides a basis for the rational design of fume hoods.

    Table and Figures | Reference | Related Articles | Metrics
    Lightweight face super-resolution network based on asymmetric U-pyramid reconstruction
    Tongyu JIANG, Fan CHEN, Hongjie HE
    Journal of Shandong University(Engineering Science)    2022, 52 (1): 1-8.   DOI: 10.6040/j.issn.1672-3961.0.2021.312
    Abstract794)   HTML18)    PDF(pc) (9804KB)(295)       Save

    A lightweight face super-resolution network was proposed in order to solve the problem that the model of deep convolutional neural network was complicated and difficult to be applied in the face super-resolution task. The coder composed of rescoding blocks was used for feature extraction, and pyramid reconstruction was introduced into the decoder to achieve fast and accurate super-resolution. To reduce the parameter number of the up-sampling operation in the decoding block, a non-uniform channel widening strategy based on resolution selection was adopted. To avoid adding extra branches, the prior knowledge of the face was introduced through heatmap loss. Experimental results showed that the model proposed in this paper could achieve light and accurate super-resolution reconstruction of ultra-low resolution face images that achieved better visual quality than the state-of-the-art method with lower model complexity.

    Table and Figures | Reference | Related Articles | Metrics
    Survey on student academic performance prediction from the perspective of task granularity
    Xiushan NIE,Yuling MA,Huiyan QIAO,Jie GUO,Chaoran CUI,Zhiyun YU,Xingbo LIU,Yilong YIN
    Journal of Shandong University(Engineering Science)    2022, 52 (2): 1-14.   DOI: 10.6040/j.issn.1672-3961.0.2021.489
    Abstract878)   HTML33)    PDF(pc) (1393KB)(294)       Save

    As one of the important research branches in educational data mining domain, student performance prediction was intensively studied. However, a comprehensive review of student performance prediction was still underexplored from the perspective of real applications. This paper detailed the technologies and methods exploited in student performance prediction research from the perspective of task granularity, and then introduced several application-oriented cases of student performance prediction, so as to provide targeted reference information for scientific researchers and educators.

    Table and Figures | Reference | Related Articles | Metrics
    Resilience evaluation system of the old community based on cloud model
    Yujiang FAN,Huanhuan HUANG,Jiaxiong DING,Kai LIAO,Binshan YU
    Journal of Shandong University(Engineering Science)    2023, 53 (5): 1-9, 19.   DOI: 10.6040/j.issn.1672-3961.0.2022.357
    Abstract202)   HTML9)    PDF(pc) (3861KB)(288)       Save

    In order to accurately and efficiently improve the living environment quality of old urban communities and improve their disaster prevention and mitigation capabilities, a resilience evaluation system of the old community based on cloud model was proposed. Based on the resilience theory, the selected old communities were grouped according to different ages. Through field research, 30 groups of representative old communities characteristic data in Xi′an, Shaanxi Province were obtained. With reference to relevant literature and relevant expert suggestions, the building resilience, facility resilience, environmental resilience, and personnel resilience were determined. A total of 4 first-level indicators, 9 second-level indicators, and 30 third-level indicators constituted the evaluation index system of old communities. Analytic hierarchy process (AHP) was used to determine the subjective weight of each index, entropy weight method (EWN) and method based on the removal effects of criteria (MEREC) were used to determine its objective weight, combined weighting method was used to determine its final weight, and MATLAB was used to construct the resilience evaluation system of the old community based on cloud model. Based on this evaluation system, an old community—Wanqingxiang community in Xi′an was selected for resilience evaluation. The results could accurately reflect the weak links of resilience in the community, and also showed that the evaluation system had certain applicability and effectiveness.

    Table and Figures | Reference | Related Articles | Metrics
    Research review of highway differentiated toll collection
    Jianqing WU,Yanqiang HUO,Jianzhu WANG,Hongyu GUO
    Journal of Shandong University(Engineering Science)    2023, 53 (4): 18-29.   DOI: 10.6040/j.issn.1672-3961.0.2023.063
    Abstract386)   HTML5)    PDF(pc) (1386KB)(280)       Save

    In order to formulate a scientific and reasonable scheme for highway differentiated toll collection, the background, realization manners, related theories and key technologies are systematically described, and the cases upgraded in Guangxi, Tianjin and Hebei are briefly introduced with design essentials and application effects, and outlooks on the research trend of highway differentiated toll collection are given.

    Table and Figures | Reference | Related Articles | Metrics
    MIRGAN: a medical image report generation model based on GAN
    Junsan ZHANG,Qiaoqiao CHENG,Yao WAN,Jie ZHU,Shidong ZHANG
    Journal of Shandong University(Engineering Science)    2021, 51 (2): 9-18.   DOI: 10.6040/j.issn.1672-3961.0.2020.227
    Abstract1064)   HTML51)    PDF(pc) (2295KB)(277)       Save

    The medical image report generation task based on image understanding became a widely concerned issue. Compared with the traditional image understanding task, medical image report generation was a more challenging task. We proposed a medical image report generative adversarial network (MIRGAN) model for this task. A co-attention mechanism was adopted to synthesize the visual and semantic features of multiple feature areas and generate descriptions corresponding to these areas. Combining the generative adversarial networks (GAN) and reinforcement learning (RL) optimized the performance of the generative model to output higher quality reports. The experiment results demonstrated the effectiveness of our proposed MIRGAN model.

    Table and Figures | Reference | Related Articles | Metrics
    Adaptive multi-domain sentiment analysis based on knowledge distillation
    YANG Xiuyuan, PENG Tao, YANG Liang, LIN Hongfei
    Journal of Shandong University(Engineering Science)    2021, 51 (3): 15-21.   DOI: 10.6040/j.issn.1672-3961.0.2020.249
    Abstract932)      PDF(pc) (2317KB)(274)       Save
    An adaptive multi-domain knowledge distillation framework was proposed, which effectively accelerated reasoning and reduced model parameters while ensuring model performance. The knowledge distillation method was used to study sentiment analysis problems. When performing knowledge distillation for each specific field, model distillation involved word embedding layer distillation, coding layer distillation(attention distillation, hidden state distillation), output prediction layer distillation and other aspects of distillation, in order to learn all aspects knowledge from the specific field teacher model. Selectively learning the importance of the teacher model corresponding to different fields to the data was proposed, which further improved the accuracy of the prediction results. The experimental results on multiple public datasets showed that after single-domain knowledge distillation increased the model accuracy by an average of 2.39%, while multi-domain knowledge distillation increased the model accuracy by an average of 0.5%. Compared with the knowledge distillation of a single domain, this framework enhanced the generalization ability of the student model and improved the performance.
    Reference | Related Articles | Metrics
    Evacuation simulation model based on multi-target driven artificial bee colony algorithm
    Xinlu ZONG,Jiayuan DU
    Journal of Shandong University(Engineering Science)    2021, 51 (3): 1-6.   DOI: 10.6040/j.issn.1672-3961.0.2020.259
    Abstract925)   HTML534)    PDF(pc) (4584KB)(273)       Save

    An evacuation simulation model based on multi-target driven artificial bee colony algorithm was presented. Based on the artificial bee colony algorithm, the visual field was used for each following bee to choose the individual with the best fitness value in the field as its visual leading bee and avoid blind choice. A multi-target driven artificial bee colony algorithm was proposed. The moving direction of each following bee was affected by multiple targets, including inertial leading bee, global optimal bee, historical optimal bee and visual leading bee. The experimental results showed that the multi-target driven artificial bee colony algorithm had higher efficiency and achieved better performance and more reasonable distribution in the case of larger number of evacuees. The model and algorithm presented could effectively improve evacuation efficiency and was suitable for the evacuation problem in multi-obstacle situation.

    Table and Figures | Reference | Related Articles | Metrics
    Short-term wind power prediction based on CEEMDAN-GRA-PCC-ATCN
    Xinzhang WU,Xiangyu LIANG,Hongyu ZHU,Dongdong ZHANG
    Journal of Shandong University(Engineering Science)    2022, 52 (6): 146-156.   DOI: 10.6040/j.issn.1672-3961.0.2022.242
    Abstract621)   HTML3)    PDF(pc) (6481KB)(273)       Save

    To improve the accuracy of wind power prediction, a short-term wind power prediction method based on data decomposition and input variable selection was proposed. The complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) was used to decompose the original wind power and wind speed data, and smooth data fluctuation to extract internal hidden information. The wind power components were simplified and reconstructed by permutation entropy (PE) algorithm to reduce the model complexity. To enhance the correlation between the input variables and wind power, eliminate redundant information and reduce the dimensionality of the input data, the Pearson correlation coefficient (PCC) and gray relation analysis (GRA) were combined to select the input variables for each reconstructed wind power component. The attention-based temporal convolutional network was used to predict the reconstructed power components, and the predicted values were superimposed to obtain the final result. The experimental results showed that the short-term wind power prediction method based on CEEMDAN-GRA-PCC-ATCN could extract more internal key information of wind power data, reduce the dimension of input data, strengthen the correlation between input variables and wind power, and effectively improve the prediction accuracy.

    Table and Figures | Reference | Related Articles | Metrics
    Intelligent scheduling technology of highway emergency rescue vehicle
    Xiuguang SONG,Xinming GUO,Fang YAN,Guoqiang LI,Yuan TIAN
    Journal of Shandong University(Engineering Science)    2023, 53 (4): 1-17.   DOI: 10.6040/j.issn.1672-3961.0.2023.045
    Abstract390)   HTML29)    PDF(pc) (2239KB)(273)       Save

    Traffic accidents had suddenness, complexity and severity. Efficient and reasonable dispatch of emergency rescue vehicles played a vital role in reducing casualties and property losses. The influencing factors of rescue route planning were introduced. The model was described from the single objective model and the multi-objective model. Combined with the characteristics of emergency rescue in sudden traffic accidents, the route optimization algorithm of emergency rescue vehicles was summarized from two aspects: precise algorithm and meta-heuristic algorithm. The current emergency rescue vehicle scheduling technology and development trend were summarized and prospected.

    Table and Figures | Reference | Related Articles | Metrics
    Prediction method of power grid emergency supplies under meteorological disasters
    Qingfa CHAI,Shoujing SUN,Jifu QIU,Ming CHEN,Zhen WEI,Wei CONG
    Journal of Shandong University(Engineering Science)    2021, 51 (3): 76-83.   DOI: 10.6040/j.issn.1672-3961.0.2020.527
    Abstract936)   HTML138)    PDF(pc) (2767KB)(261)       Save

    In order to improve the efficiency of grid emergency repairs, a method for predicting emergency supplies under the conditions of power grid meteorological disasters combining case-based reasoning, deep belief network and deep learning was proposed. Based on meteorological data, power grid maintenance data and geographic environment data, this method was used case-based reasoning to determine the appropriate input and output structure of the prediction model, and different methods was used to process and quantify according to the characteristics of disagreeing input factors. Deep belief networks were used to complete case adaptation, and integrate accident scale information was used to establish a dynamic power grid emergency supplies prediction model. The analysis results showed that the emergency material prediction method proposed in this paper could comprehensively analyze various characteristic factors, and combined the scale of the accident to establish the relationship between the emergency material demand of the power grid, and accurately predicted the material demand for the emergency response of the power grid under the weather disaster. and provided a scientific reference for emergency decision-making of power grids.

    Table and Figures | Reference | Related Articles | Metrics
    Experimental study on mechanical properties and pore structure of gangue mixed aggregate concrete
    LI Wen, WANG Hailong, ZHANG Jiahao, YANG Hong, WANG Lei, FENG Shuai
    Journal of Shandong University(Engineering Science)    2023, 53 (3): 121-127.   DOI: 10.6040/j.issn.1672-3961.0.2021.449
    Abstract211)      PDF(pc) (3676KB)(231)       Save
    Reference | Related Articles | Metrics
    The classification of mild cognitive impairment based on supervised graph regularization and information fusion
    Ying LI,Jiankun WANG
    Journal of Shandong University(Engineering Science)    2023, 53 (4): 65-73.   DOI: 10.6040/j.issn.1672-3961.0.2023.025
    Abstract210)   HTML4)    PDF(pc) (3086KB)(231)       Save

    To precisely distinguish progressive and stable mild cognitive impairment (MCI). The projection matrix was learned from Alzheimer′s disease samples and normal control samples. The supervised graph regularization was used to optimize the local nearest neighbor relationship of the samples. Based on the projection matrix, the spatial transformation of the MCI samples was carried out to extract the discriminative features of progressive and stable MCI. The proposed features were fused with the scores of Mini-Mental State Examination and apolipoprotein E4. The SVM classifier was trained using the fused features for the MCI classification. The experiments were conduct on the Alzheimer′s Disease Neuroimaging Initiative (ADNI) database. The classification accuracy reached to 73.33%. Compared with the existing approaches, the proposed method significantly improved the classification accuracy, sensitivity and specificity.

    Table and Figures | Reference | Related Articles | Metrics
    Reduced-order analytical model to evaluate photovoltaic low-voltage ride-through performance
    NIU Shuanbao, HUO Chao, CHEN Chunmeng, KE Xianbo, WANG Xiaohui, ZHANG Qiang, CHEN Ning
    Journal of Shandong University(Engineering Science)    2021, 51 (3): 91-100.   DOI: 10.6040/j.issn.1672-3961.0.2020.472
    Abstract755)      PDF(pc) (3951KB)(226)       Save
    A reduced-order analytical model to evaluate the performance of PV in case of low-voltage ride through(LVRT)was proposed based on the idea of order reduction at each unit, focusing on the DC bus voltage which determined the outcome of ride-through. Such a method covered the major factors with regard to PV LVRT, and included the responses by the converter controller and phase-locked loop. Meanwhile, it took the supporting of reactive power outputs from PV on the point of common coupling(PCC)voltage into consideration, so as to reflect that PV provides reactive power support to AC system. To validate the proposed reduced-order analytical model, the comparison with the simulation results derived from PSCAD confirmed its accuracy and feasibility.
    Reference | Related Articles | Metrics
    Intelligent commutation system of distribution network based on differential evolution algorithm
    Yunlong ZHAO, Renfei CHE, Jiahui CHEN
    Journal of Shandong University(Engineering Science)    2021, 51 (5): 107-113.   DOI: 10.6040/j.issn.1672-3961.0.2020.110
    Abstract749)   HTML24)    PDF(pc) (1771KB)(214)       Save

    The application of three-phase intelligent commutation switch was an effective measure to solve the problem of three-phase load imbalance in low-voltage distribution network. Based on the commutation process of the three-phase intelligent commutation switch, an intelligent commutation system of the distribution network was proposed, which took the differential evolution algorithm as the core and the mixed solid-state switch as the foundation. Aiming at the unbalance degree of three-phase current and the action times of commutation switch, differential evolution algorithm was used to optimize the commutation strategy. The optimal commutation strategy was obtained through initialization, mutation, crossover and selection operation, which improved the real-time and accuracy of the system. In this paper, through the combination of magnetic holding relay and IGBT based solid-state switch, a fast switching method using hybrid solid-state switch was proposed, which the commutation switch showed the characteristics of low power consumption and strong current passing ability, and improved the reliability of the system. The feasibility of the algorithm and switch in fast switching and im-proving the power quality of the commutation process was verified by practical test.

    Table and Figures | Reference | Related Articles | Metrics
    Mechanical characteristics of TBM tunnel segment in composite stratum
    Weiqin ZHENG,Jie XU,Jie SUN,Ke WU
    Journal of Shandong University(Engineering Science)    2022, 52 (4): 210-213.   DOI: 10.6040/j.issn.1672-3961.0.2021.259
    Abstract670)   HTML4)    PDF(pc) (4433KB)(210)       Save

    Based on the redistribution of soil stress around the excavated tunnel due to the large buried depth of the tunnel during the construction of the composite stratum by the full section hard rock tunnel boring machine (TBM), considering that the surrounding rocks would produce the force of squeezing into the tunnel, segment damage might occur in the tunnel construction. The large-scale finite element numerical software ANSYS and finite difference software FLAC3D were used to simulate the underpass formation. The stress and strain of tunnel construction segments were comprehensively analyzed, and reasonable monitoring methods and protection measures were put forward. The main conclusions were as follows, when only considering the influence of stratum stress redistribution on lining segments after double tunnel construction, it was suggested that double tunnels should be constructed at the same time, and the displacement of segments was less affected by the direction of tunnel construction; In the construction process, the two adjacent sides of the double tunnel segment would have large horizontal displacement. If necessary, the inner side of the segment should be monitored and reinforced.

    Table and Figures | Reference | Related Articles | Metrics
    Wind induce vibration control of chimney with suspended flue
    LI Xu, LIU Bing, CHEN Ying, WANG Peijun
    Journal of Shandong University(Engineering Science)    2021, 51 (3): 101-112.   DOI: 10.6040/j.issn.1672-3961.0.2020.281
    Abstract715)      PDF(pc) (4454KB)(202)       Save
    The wind induced vibration control of a high rise reinforced concrete chimney with suspended flues(CSMF)was conducted by using rotational friction dampers(RFD). The detailing of a RFD restraint joint was presented. The wind induced vibration control effect was firstly investigated, and the state space formulation of the equation of motion of CSMF was derived based on Lagrange's equations. A CSMF of 175 m high was then taken an example to investigate the feasibility of the proposed control strategy. Five vibration control schemes were compared and the results showed that the vibration of CSMF could be substantially suppressed by using RFD with optimal parameters, position and suspended flue length. FRD was superior to TMD due to its small auxiliary mass and little space occupation.
    Related Articles | Metrics
    Internal exposure levels and influencing factors of serum per- and polyfluoroalkyl substances in the community residents in Jinan City
    ZHU Yuting, YAN Wenbao, XU Fei, JIANG Zhengting, ZHAO Nan, DING Lei
    Journal of Shandong University(Engineering Science)    2023, 53 (3): 147-154.   DOI: 10.6040/j.issn.1672-3961.0.2022.398
    Abstract280)      PDF(pc) (927KB)(200)       Save
    Reference | Related Articles | Metrics
    The analysis of key parameters of hydraulic energy storage system of wave energy converter
    Yingxin LIU,Jian QIN,Yanjun LIU
    Journal of Shandong University(Engineering Science)    2021, 51 (6): 1-8.   DOI: 10.6040/j.issn.1672-3961.0.2020.429
    Abstract957)   HTML40)    PDF(pc) (4298KB)(199)       Save

    To improve the efficiency of wave power generation technology, a mathematical model of the system from wave input to motor output was established. The key parameters that affected the system′s power generation capacity were analyzed by using theoretical modeling and simulating, so as to provide theoretical guidance for the research of the constant speed control strategy of the hydraulic energy conversion system. The hydraulic energy storage system of wave energy generation was composed of 3 parts. The mathematical model of the system was established by analyzing each component′s motion equation and energy equation, and finding the connection parameters between the two components. The key parameters and characteristics of the system were determined qualitatively by analyzing the system′s power equation. To confirm the accuracy of theoretical analysis, the AMEsim simulation platform was used to design and imitate the system. The results showed that the motor′s output power was affected by the height of the wave, period, flow area of the proportional flow valve and the motor displacement. The highest order of the influence was 1, 4, 4 and 2, respectively. The results also verified that the precharging pressure of the accumulator had little influence on the motor′s output power.

    Table and Figures | Reference | Related Articles | Metrics
    Cross social network user alignment via fusing node state information
    Jun HU,Dongmei YANG,Li LIU,Fujin ZHONG
    Journal of Shandong University(Engineering Science)    2021, 51 (6): 49-58.   DOI: 10.6040/j.issn.1672-3961.0.2021.281
    Abstract962)   HTML67)    PDF(pc) (3630KB)(198)       Save

    A cross social network user alignment method by fusing node state information was proposed. The local characteristics of nodes and node state information were captured through network representation to obtain the embedded vector of each account, and the aligned users were found by calculating the similarity between corresponding representations of different accounts. Experimental results on two real data sets showed that the proposed method could align more users than other methods. When predicting top-k of different scales, the proposed method could achieve an alignment precision of 50% at top-9 on the data set Twitter-Foursquare with dense network structure. Compared with other methods on the sparse and large network data set DM-ML, the improvement on alignment precision was 12.06%-36.62%. The analysis of F1-score also showed that the proposed method could effectively improve the performance of user alignment.

    Table and Figures | Reference | Related Articles | Metrics
    Vibration caused by P-wave incident on a rock slope with a bedding structural plane
    CHAI Shaobo, SHI Jiehui, ABI Erdi, LIU Qin, SONG Lang
    Journal of Shandong University(Engineering Science)    2023, 53 (3): 31-40.   DOI: 10.6040/j.issn.1672-3961.0.2021.506
    Abstract195)      PDF(pc) (3229KB)(195)       Save
    Reference | Related Articles | Metrics
    Policy analysis and countermeasures for achieving carbon peak and carbon neutrality
    Xueliang YUAN,Yue YANG,Xuerou SHENG,Leping CHEN,Xin DAI,Qingsong WANG,Qiao MA
    Journal of Shandong University(Engineering Science)    2023, 53 (5): 132-141.   DOI: 10.6040/j.issn.1672-3961.0.2022.406
    Abstract236)   HTML10)    PDF(pc) (2999KB)(194)       Save

    To clarify the policy system of China′s carbon peaking and carbon neutrality goals, literature in this field from 2011 to the present was quantitatively analyzed, numbers of national policies and provincial policies since China first proposed the carbon peak and carbon neutrality goals on September 22, 2020 were systematically sorted out, and the development trend of the carbon peaking and carbon neutrality policies were analyzed. Policy recommendations were proposed in the areas such as improving laws and regulations, optimizing the carbon emission trading system, promoting standardization, driving technological innovation, promoting regional collaborative emission reduction, and promoting energy system transformation. The findings had reference significance for the revision and improvement of China′s carbon peaking and carbon neutrality policies, as well as the formulation of carbon emission reduction policies in other countries.

    Table and Figures | Reference | Related Articles | Metrics
    An efficient and lightweight RGB frame-level face anti-spoofing model
    Jiachun LI,Bowen LI,Jianbo CHANG
    Journal of Shandong University(Engineering Science)    2023, 53 (6): 1-7.   DOI: 10.6040/j.issn.1672-3961.0.2023.132
    Abstract119)   HTML7)    PDF(pc) (2618KB)(192)       Save

    Based on the challenge when deploying a deep learning-based face anti-spoofing (FAS) algorithm on general-purpose consumer devices with only RGB camera, an efficient and lightweight RGB frame-level FAS model (EL-FAS) was proposed. To improve the generalization ability of the model and achieve high performance under constrained conditions, a novel global spatial self-attention mechanism was explored to capture global feature dependencies, and an equal-channel pixel-wise binary supervision method was designed to force our model to learn shared features from different pixels. The Bottlenecks residual block was used to establish the backbone network to reduce parameters. Analysis and the experimental results showed that EL-FAS model achieved state-of-the-art performance in the OULU-NPU dataset, obtained competitive performance in the SiW dataset and cross-dataset tests. The model was lightweight, with only 1.34×106 parameters.

    Table and Figures | Reference | Related Articles | Metrics
    Short-term load forecasting of iron and steel industry area based on combination model of SVM and LSTM
    Xiaoyan QI,Hengjie LIU,Qiuhua HOU,Xiaoyu LIU,Yanchao TAN,Liancheng WANG
    Journal of Shandong University(Engineering Science)    2021, 51 (4): 91-98.   DOI: 10.6040/j.issn.1672-3961.0.2020.539
    Abstract946)   HTML110)    PDF(pc) (5138KB)(186)       Save

    A short-term load forecasting algorithm combining long short-term memory (LSTM) and support vector machine (SVM) was proposed to solve the low accuracy problem of short-term load forecasting due to the large-scale iron and steel enterprise power consumption impact on regional load. The research thoroughly analyzed the load characteristics of the selected region with predominant iron and steel mill load, which divided the load into the impulse load and others based on its various components.Covariance algorithm and Pearson algorithm were used to analyze the correlation and differentiation of load influence factors. Six attributes of historical load, temperature, date type, steel price, electricity price and iron ore price were selected as load forecasting. The fuzzy weight assignment was used to fuse LSTM and SVM which got the final load forecasting result. The simulation results showed that the proposed method could predict the short-term load more accurately than the single LSTM or SVM.

    Table and Figures | Reference | Related Articles | Metrics
    Median calculation algorithms based on GPU in OLAP
    WU Zhenpeng, ZHANG Jian, FAN Xingqi, LI Cuiping
    Journal of Shandong University(Engineering Science)    2021, 51 (3): 7-14.   DOI: 10.6040/j.issn.1672-3961.0.2019.508
    Abstract944)      PDF(pc) (3119KB)(185)       Save
    An algorithm for one of the holistic aggregate operations in online analytical processing(OLAP)called Median was proposed based on graphics processing unit(GPU), which was named GPU-Median algorithm. This algorithm obtained the median of a series of data by segmenting the data, sorting the data by segments,cutting the data preceding the global median, and finally merging the uncut data. Through the algorithm above much time spent on global sorting was saved. Then an algorithm called GPU-Median+was presented in order to optimize and extend the GPU-Median algorithm. This algorithm implemented the aggregate operations through the collaboration of CPU and GPU, which used GPU to deal with segments of data and CPU to deal with global data. Experiments and analysis proved that the GPU-Median + algorithm reduced the time complexity of the median calculation from O(n2)to O(n)compared to the CPU algorithm,and that the GPU-Median + algorithm reduced a third of the calculation time compared to the radix sort algorithm on the GPU. The application of this algorithm enabled the GPU to improve its ability of parallel calculations when calculating the holistic aggregate function in OLAP, thus providing a new idea for improving the query performance of OLAP.
    Reference | Related Articles | Metrics
    Review of frequency dynamic behavior evolution and analysis method requirements of power system
    Hengxu ZHANG,Yongji CAO,Yi ZHANG,Changgang LI,Jiacheng RUAN,VLADIMIR Terzija
    Journal of Shandong University(Engineering Science)    2021, 51 (5): 42-52.   DOI: 10.6040/j.issn.1672-3961.0.2021.174
    Abstract926)   HTML32)    PDF(pc) (1810KB)(180)       Save

    The development of renewable energies changed the operation states of power systems, which complicated the frequency dynamics and brought about new challenges to frequency security and stability. The process of frequency dynamic response was introduced, whose quantitative characteristics and analysis methods were reviewed. And the strengths and promising prospects of frequency dynamics analysis method based on artificial intelligence were emphasized. From the viewpoints of the strong variability of renewables energies, the low inertia of nonsynchronous power sources, and the high risk of large power deficits and cascading faults, the changes of the operation states of power systems were analyzed, and the new requirements on frequency analysis and control were discussed. The inertia definition of power systems with the rapid development of renewable energies was revealed, and the virtual inertia control methods were reviewed. The strengths and promising prospects of the coordination control of multi-type virtual inertia were emphasized. The important issues to be focused in the research area of frequency analysis and control were summarized, and the suggestions for further studies were provided.

    Table and Figures | Reference | Related Articles | Metrics
    Fault diagnosis of vehicle-to-vehicle communication in networked traffic environment
    Haigen MIN,Yukun FANG,Xia WU,Wuqi WANG
    Journal of Shandong University(Engineering Science)    2021, 51 (6): 84-92.   DOI: 10.6040/j.issn.1672-3961.0.2021.353
    Abstract678)   HTML28)    PDF(pc) (7549KB)(179)       Save

    With vehicle platoon in intelligent transportation system as the background, this research established a vehicle platoon control model based on the intelligent driver model and analyzed the influence of delay in vehicle-to-vehicle communication on the stability of platoon control. A fault diagnosis method based on the update delay of vehicle-to-vehicle communication was proposed. The statistical characteristics of median and average were used to calculate the decision variables which were used to judge whether a fault had occurred. A two-layer sliding window was designed to smooth the decision variables and adaptively calculate the decision variable in real time. The Jarque-Bera algorithm was used to test the normality of the statistical distribution of the receiver's update delay within a period of time. If the distribution significantly deviated from the normal distribution, it was considered that the communication quality had deteriorated. Collecting vehicle speed data and vehicle-to-vehicle communication delay data at the testbed, simulations were conducted to analyze the statistical distribution characteristics of the vehicle-vehicle communication update delay in different scenarios and verify the influence on vehicle platoon control of vehicle-to-vehicle communication delay. The research results showed that the vehicle-to-vehicle communication delay caused drastic changes in the control rate during the collaborative control process. The communication fault diagnosis method based on the update delay could effectively diagnose whether the vehicle-to-vehicle communication quality had deteriorated.

    Table and Figures | Reference | Related Articles | Metrics
    An intrusion detection model based on improved ReliefF algorithm
    Caihui LIU,Qi ZHOU,Xiaowen YE
    Journal of Shandong University(Engineering Science)    2023, 53 (2): 1-10.   DOI: 10.6040/j.issn.1672-3961.0.2022.136
    Abstract386)   HTML20)    PDF(pc) (3888KB)(171)       Save

    Aiming at the problems of insufficient feature extraction in the existing intrusion detection algorithms, the influence of feature weights was not considered, and the model classification was not accurate enough, an intrusion detection model based on the improved ReliefF algorithm was proposed. By optimizing the calculation of the feature weight of the intrusion data, an improved algorithm of ReliefF was proposed, based on the Pearson correlation coefficient of the calculated feature, a feature correlation scale was established. Only one of the features with high correlation was retained to realize the secondary optimization of the features, and finally decision tree, k-nearest neighbor, random forest, naive bayes and support vector machine classifier were used to evaluate the classification performance and accuracy. Experimental results on NSL-KDD and UNSW-NB15 data sets showed that this method could not only effectively reduce the feature dimension, but also had better detection performance, which had a positive effect on the computational complexity of the classifier.

    Table and Figures | Reference | Related Articles | Metrics