Table of Content

    25 March 2020, Volume 59 Issue 2
     A concise survey on graph convolutional networks
    2020, 59(2):  1-14.  doi:10.13471/j.cnki.acta.snus.2020.02.001
    Asbtract ( 174 )   PDF (2215KB) ( 167 )  
    Related Articles | Metrics
    In recent years, many new technologies are constantly emerging in every aspect of our lives. More and more data have been generated and stored in graph format. Graphs are irregular data, which possess the characteristic of being distributive and disordered. Besides its capability that nodes can endow with data features, edge information can further depict the similarities among nodes. Despite the fact that classic convolutional neural networks are capable of handling regular format data such as images, videos and speech, directly applying these networks to graph data seems to be problematic. Recently, quite a few of researches were proposed to consider how to generalize classic convolutional neural networks for graph data and many high efficient learning algorithms were developed. This work aims to summarize and discuss the promising development of graph convolutional neural networks that were specifically designed for graph data. Nonetheless, due to the limited space, we cannot provide all the details of graph convolutional neural networks. Instead, we tend to introduce the motivations of those models, the analyses of the pros and cons of each model, and a brief summary of the major applications of graph convolutional neural networks.
    A class of Diophantine equation
    DOU Xiaoxia
    2020, 59(2):  15-21.  doi:10.13471/j.cnki.acta.snus.2020.02.002
    Asbtract ( 65 )   PDF (878KB) ( 43 )  
    Related Articles | Metrics
    A class of Diophantine equation ax2+by2+cz2=m+dxyz was considered. Based on the definition of the fundamental solution, with the help of the binary quadratic forms and the fundemental mathmatical knowledge, all of the possibilities of positive integers a, b, c, d and nonnegative integer m in the equation could be figured out when ald, bld and cld. And meanwhile, the corresponding fundamental solution was computed in each case. By these fundamental solutions, many integer solutions of every equation can be derived.
    Weyl's theorem for bounded linear operators and its functional calculus
    2020, 59(2):  22-27.  doi:10.13471/j.cnki.acta.snus.2020.02.003
    Asbtract ( 52 )   PDF (885KB) ( 30 )  
    Related Articles | Metrics
    Let B(H) be the set of all bounded linear operators on an infinite dimensional separable complex Hilbert space H. If T∈B(H), the set of all functions f which are analytic on a neighbourhood of σ(T) and are not constants on any component of σ(T) is denoted by H(T). Using the new spectrum, Weyls theorem for T∈B(H) and f(T) (f∈H(T)) is studied, and the necessary and sufficient conditions for both T and f(T) satisfying Weyl's theorem are characterized. Moreover, Weyls theorem for p-hyponormal (or M-hyponormal) operators are explored.

    Stability and Hopf bifurcation of a tumor immune model with time delay
    LIANG Yuqin, JIA Yunfeng
    2020, 59(2):  28-33.  doi:10.13471/j.cnki.acta.snus.2020.02.004
    Asbtract ( 53 )   PDF (868KB) ( 23 )  
    Related Articles | Metrics
    A tumor immune model with time delay and diffusion term is considered. The stability of the boundary and positive equilibrium states and the existence of the Hopf bifurcation emitting from the positive steady-state are studied. The results show that the time delay can destroy the stability of the positive equilibrium state and lead to the occurrence of the Hopf branch. 
    Research on dynamic variable evaluation model of geological risk assessment for rail transit 
    WU Zhiyi, GAO Yan, YUAN Quan, LI Wenlong
    2020, 59(2):  34-47.  doi:10.13471/j.cnki.acta.snus.2020.02.005
    Asbtract ( 67 )   PDF (1541KB) ( 56 )  
    Related Articles | Metrics
    As the lifeline of urban traffic, the health evaluation of rail transit has become increasingly important. In recent years, the models to estimate the health of tunnel engineering was mainly concentrated on highway tunnels, mountain tunnels or the operation stage of rail transit tunnel. It is still lack of models to assess the geological risk in the design and construction stages of rail transit. Based on a large number of literatures and standards, a risk criteria including 11 geological risk sources and 40 identification indexes was set up, in which four risk levels, i.e., small, weak, medium and strong, was utilized to evaluate the overall geological risk. The weights of each index were determined by BP neural network and more flexible multiplication scale method. Then, a fuzzy variable evaluation model for geological risk in rail transit tunnel was established. The time factor set was also added into the fuzzy variable evaluation model to realize the dynamic evaluation of geological risk which can help to estimate the risk during the whole life of rail transit. Finally, the model was proved by the calculating results based on the preliminary geological data of a proposed rail transit line in Guangzhou. The risk levels evaluated by the model were consistent with the insitu measurement and geological survey.

    Shared bicycle regional allocation planning model based on bi-level programming
    HU Yucong, LI Junting, CHEN Zhiwei
    2020, 59(2):  48-59.  doi:10.13471/j.cnki.acta.snus.2020.02.006
    Asbtract ( 74 )   PDF (2301KB) ( 21 )  
    Related Articles | Metrics
    In recent years, bicycle sharing system has gradually become one of the main ways of short distance travel for residents, but the problem of unbalanced between the allocation of bicycles and user's need remains to be solved. In order to study the location and allocation of shared bicycles, the concept of shared bicycle regional allocation and virtual station is presented, a shared bicycle network is built, and also a shared bicycle regional allocation planning model which based on bi-level programming is proposed. The upper model aims at minimizing the total travel cost of users, and considers the constraints of enterprises and users, which is the optimal model of the system, while the lower part is a system user equilibrium model. A hybrid genetic algorithm and particle swarm optimization algorithm is proposed to solve this model. Finally, a case study is conducted on the background of shared bicycle travel data in a mediumsized city, to verified the effectiveness of the model and algorithm in the actual. The result shows that this model can determine the location and allocation of the virtual stations in each region, thus can provide a scientific and reasonable basis for the deployment of sharing bicycles.

    Research on finite element model updating multi-alternatives problem based on niche fish swarm algorithm
    KAGN Juntao, ZHANG Xueqiang, ZHANG Yazhou, QIN Shiqiang, CAO Hongyou, LIAX Yuequan
    2020, 59(2):  60-68.  doi:10.13471/j.cnki.acta.snus.2020.02.007
    Asbtract ( 60 )   PDF (1633KB) ( 36 )  
    Related Articles | Metrics
    The traditional model updating technique uses intelligent algorithms to find a set of global optimal solutions,which makes the finite element model closer to the actual structural response. Actually,since the number of objective functions is often smaller than the number of parameters to be optimized,there may be more than one global optimal solution. Because the number of sensors is limited,the measurement data is insufficient,and the instrument identification has errors,the local optimal solution may represent the actual structural parameters more than the global optimal solution. Based on that,the paper proposes a niche fish swarm algorithm,which introduces niche technology and reverse learning strategy,and uses simulated annealing algorithm to perform detailed optimization in the later stage of the algorithm,It overcomes the drawbacks of the artificial fish swarm algorithm that it is difficult to find multiple function peaks,sensitive to initial values,slow convergence and low precision,and it is verified by numerical simulation to apply it to a scaled cablestayed bridge model. In order to avoid a large number of calls to the finite element model,this paper uses the response surface as a proxy model,generates experimental data through the Latin hypercube,performs a significant test on each factor,and eliminates the unrelated factors to test the fitting accuracy. The model updating results in a set of global optimal solutions and 7 sets of local optimal solutions that reduce the error between the calculated and measured values of the scaled cable-stayed bridge,Thus, decision makers can choose from multiple sets of solutions based on engineering experience, reducing the possibility of misjudgement of actual structural parameters.

    Research on automatic control algorithm for central tire inflation system of heavy-duty vehicle
    XU Jun, SU Tian, LI Gangyan, CHENG Yi
    2020, 59(2):  69-76.  doi:10.13471/j.cnki.acta.snus.2020.02.008
    Asbtract ( 56 )   PDF (3557KB) ( 12 )  
    Related Articles | Metrics
    As a key load-bearing part of heavyduty vehicle, the tire pressure should be supervised in real time and be adjusted through central tire inflation system to ensure safety driving. The common control algorithm based on pressure difference with calculation time will cause the actual value of tire pressure deviates greatly from the set value without considering the influence of dynamic pressure during tire inflation and deflation process. Based on the analysis of tire pressure characteristics, a fuzzy  compensation control algorithm which integrates checking and scheduling of inflation & deflation function was proposed to implement accurate adjusting tire pressure through fuzzy fast inflation and error compensation. Taking a 6×4 heavyduty vehicle as the application object, an AMESim-MATLAB co-simulation platform was set up to simulate and verify the control algorithm. The acquired tire pressure through CAN bus was taken as the input parameter for the central tire inflation system. Through simulation of fault injection test and loading verification, the test results have shown that the controller prototype based on the control algorithm can realize automatic control of the tire pressure of heavyduty vehicle, and the fault response time is less than the national recommended standard requirements.

    Seismic damage investigation and vulnerability analysis of masonry structures in different intensity zones
    LI Siqi, YU Tianlai, ZHANG Ming
    2020, 59(2):  77-85.  doi:10.13471/j.cnki.acta.snus.2020.02.009
    Asbtract ( 58 )   PDF (2413KB) ( 29 )  
    Related Articles | Metrics
    This paper analyses the seismic damage of 3 991 masonry structures in Dujiangyan during Wenchuan earthquake, to understand the seismic damage characteristics and seismic capacity of masonry structures in different intensity regions. Firstly, seismic damage of Dujiangyan masonry structure is observed. The results display that the location of this kind of structure system is easy to be damaged: vertical and horizontal walls, openings, joints of vertical and horizontal walls, window (lower) walls and tie columns; However, a large number of masonry structures are basically intact after seismic design. Using the method of numerical statistical analysis to analyze the survey data, the relationship between the sample and the seismic damage grade and the vulnerability matrix are given, and the nonlinear model of the seismic damage grade and vulnerability under different intensity grades in the multi intensity region is established. Considering the seismic damage of the structure under the influence of seismic fortification factors, the general law of seismic damage index and intensity of the structure in multiintensity region is analyzed by regression curve. The results can be used for reference in the revision of intensity scale and vulnerability analysis.

    Passenger taxi boarding demand prediction via CorrelationNet with Dropconnect
    HUANG Min, MAO Feng, QIAN Yuxiang, SHA Zhiren
    2020, 59(2):  86-94.  doi:10.13471/j.cnki.acta.snus.2020.02.010
    Asbtract ( 53 )   PDF (2080KB) ( 13 )  
    Related Articles | Metrics
    To improve the accuracy of taxi customer pick-up prediction and the operation efficiency of taxi service, a CorrelationNet with dropconnect method is proposed. The method includes two stages, spatiotemporal features identification and regularization with dropconnect. First, the prediction model adds the spatiotemporal correlation analysis mechanism to the deep neural network, and determines the spatiotemporal features for taxi demand prediction, forming a new deep learning network called CorrelationNet. Then, the prediction model uses dropconnect method to train the new deep learning network Correlationnet to reduce over fitting. Finally, a case study is carried out in Guangzhou to verify the model; support vector machine regression (SVR), artificial neural network (ANN) and CorrelationNet methods are adpoted to evaluate the performance of the proposed method by using the same taxi GPS data. The experiments results show that this method is better than other methods and is more suitable for passenger taxi demand prediction.

    Experimental study on fire combustion of fullscale bus
    ZHANG Liang, ZHANG Desheng, LIANG Guofu, LU Zhibao, JI Chao
    2020, 59(2):  95-100.  doi:10.13471/j.cnki.acta.snus.2020.02.011
    Asbtract ( 43 )   PDF (1809KB) ( 13 )  
    Related Articles | Metrics
    In recent years, the proportion of bus fire has increased significantly. Because of the dense staff and the lack of cognition about bus arson, it often causes mass casualties and injuries. This article conduct a bus combustion experiment, simulated the“5.9”school bus arson case in Weihai , which is ignited by arson in cab.The temperature and smoke changes of the vehicle during the combustion spread are observed, and the spread law of this kind of fire is summarized.The research results showed that the fire in the driver's cabin spreads rapidly, and the oxygen in the vehicle is rapidly consumed, generating a large amount of toxic gases such as CO, nitrogen oxides and sulfides after the bus is burned,. The temperature inside the vehicle exceeds 1 000 ℃ in about 3 minutes. The best time to escape from the bus is less than 30 s by breaking the window.

    The road-region boundary based on vegetation index
    LIANG Kaixuan, ZHANG Guifang, LI Hongzhong, WEN Jinfang
    2020, 59(2):  101-109.  doi:10.13471/j.cnki.acta.snus.2020.02.012
    Asbtract ( 58 )   PDF (3568KB) ( 30 )  
    Related Articles | Metrics
    The research on road-region boundary based on vegetation index is of great significance for road construction decision and eco-environmental protection. Taking Landsat 8 OLI remote sensing image of Guangkun highway as a case, the law of vegetation growth in different buffer zones was discussed based on vegetation index, so as to provide a criteria for determining road-region boundary. The results showed that NDVI, ARVI and IAVI vegetation indices can be used for monitoring road ecological changes constantly. The influence of road on adjacent vegetation will decrease with the increase of distance. The influence is strong within 200 m, and tends to be stable around 600 m. Moreover, road-region boundary can be scaled dynamically in which road-region boundary should be relaxed to the minor influence area or the non-influence area in route selection or reduced to the main affected area in ecoenvironmental assessment.
    Spatiotemporal patterns of the shallow groundwater depth across the Huaibei Plain 
    YAO Rui, SUN Peng, ZHANG Qiang, JIANG Shangming, XIA Min, WANG Junhong
    2020, 59(2):  110-119.  doi:10.13471/j.cnki.acta.snus.2020.02.013
    Asbtract ( 57 )   PDF (5156KB) ( 24 )  
    Related Articles | Metrics
    Using the monthly groundwater depth at 104 stations and the monthly precipitation at 97 stations across the Huaibei plain, China, covering the period of 1980-2007, trends and characteristics of groundwater depth were analyzed, and regional responses of the ground water depth to spatialtemporal distribution of precipitation were also investigated. The results indicated that: (1) The groundwater depth was the smallest in the 1980s and was the largest in the 1990s. Massive agricultural irrigation due to the extreme droughts was the major driving factor behind deepening processes of the groundwater level. (2) The groundwater level in the south was less deep than that in the north, and groundwater depth was the deepest in northeastern Huaibei plain. On a seasonal scale, the average depth of groundwater showed the smallest in summer and the largest in spring. (3) Groundwater depth increased during the growing season (from March to September) during 1980-1985 and 1990-2007, and decreased during 1985-1990. (4) Although precipitation reduced, the groundwater depth was decreasing in the central and western Huaibei plain from March to May and in September. In eastern region, the groundwater depth was decreasing from August to September, caused by decreased precipitation.

    Evolution of spatial distribution pattern and its influencing factors of talents in the Pearl River Delta
    ZHANG Jianing, LI Lixun
    2020, 59(2):  120-130.  doi:10.13471/j.cnki.acta.snus.2020.02.014
    Asbtract ( 49 )   PDF (1840KB) ( 34 )  
    Related Articles | Metrics
    Based on county-level National Demographic Census data in 2000 and 2010, the spatial auto-correlation analysis was applied to specify the spatial distribution patterns of talents in the Pearl River Delta. Multiple regression was used to analyze the main factors behind the spatial distribution patterns. Two kinds of talents were specially focused: creative talents in terms of the occupational skill and highly educated talents in terms of the educational attainment. The results indicate that: (1) There has witnessed the significant concentration among both of creative talents and highly educated talents, while the concentration trend decreased gradually. On the countylevel unit scale, talents are concentrated at Yuexiu District, Haizhu District, Tianhe District, Bao'an District, Nanshan District, Futian District, Shunde District, etc; less talents stay in counties of the east and west Pearl River Delta. (2) The distribution of both of creative talents and highly educated talents has been influenced by economic factors (e.g., regional economic development, wage level and urbanization level) and amenity (e.g.real estate, education). However, there are differences in the factors affecting the spatial distribution of talents. Education and tolerance had an impact on the distribution of highly educated talents, while finance had a significant positive impact on spatial distribution of creative talents.

    Establishment and Evaluation on the Animal Model of Qi deficiency and blood stasis syndrome in hyperlipidemia rats with exhaustive swimming exercising
    ZHANG Weijian, LIU Hong, SU Weiwei, ZENG Xuan, WANG Yonggang, LI Peibo, PENG Wei, YAO Hongliang
    2020, 59(2):  131-136.  doi:10.13471/j.cnki.acta.snus.2020.02.015
    Asbtract ( 35 )   PDF (1043KB) ( 13 )  
    Related Articles | Metrics
    The aim of the study was to establish the animal model of qi deficiency and blood stasis syndrome for drug reasearch. Rats were consecutively fed with high fat diet and then treated with exhaustive swimming exercising. The external signs of rats were observed and twenty-six biochemical parameters were detected, the parameters included energy metabolisms, oxidative stress, lipid metabolism, inflammatory cytokines, immune response, myocardial enzyme, endothelial function and clotting time etc. The results showed that the model rats were inactive, weak and appeared in lethargy, and showed a dark purple tongue, reflecting the clinical characteristics of Qi deficiency and blood stasis syndrome. Compared with the normal group, the model rats showed significant changes in the biochemical indexes mentioned above. All results together indicated that high fat diet combined with the continual exhausting swimming could cause the disorders of energy metabolisms, oxidative stress, and so on. Such disorders are closely associated with the generation and development of Qi deficiency and blood stasis syndrome. Our study demonstrated that continual exhausting swimming followed by high fat diet could be a method to make a rat model of Qi deficiency and blood stasis syndrome. The rat model could be further utilized in the evaluation of drugs for cardiovascular disease due to its good stability.

    Efficiency comparison of Rosa26 targeted integration of exogenous gene via ZFN, TALEN or CRISPR/Cas9
    LIU Xiaofeng, LIU Wei, NIE Yu, CONG Peiqing, LIU Xiaohong, CHEN Yaosheng, HE Zuyong
    2020, 59(2):  137-144.  doi:10.13471/j.cnki.acta.snus.2020.02.016
    Asbtract ( 27 )   PDF (2908KB) ( 8 )  
    Related Articles | Metrics
    Mammalian melanin synthesis is dependent on the oxidation of tyrosine. Tyrosinase (Tyr) is a key enzyme that catalyzes the oxidation of tyrosine. When exogenous Tyr  is integrated into the genome of white coat color mouse, it can render the mouse to obtain the function of melanin synthesis, presenting a different coat color. To rapidly generate the mouse model with Tyr  gene integrated, in this study, we constructed a promotorless plasmid donor pTyr-2A-DsRed, which containis the coding sequences of Tyr  gene and the red fluorescent reporter (DsRed), and the two flanking homologous arms. We selected the first intron of Rosa26 for targeting integration, and designed ZFN pair, TALEN pairs and CRISPR/Cas9 cutting at almost the same site. Through measuring the intensity of DsRed fluorescence in C2C12 cells by flow cytometry, we compared the efficiency of the targeted integration of exogenous DNA mediated by the three different genome editing tools, and found that CRISPR/Cas9 was the most efficient. Therefore, we further integrated the plasmid donor into the genome of mouse embryonic stem (ES) cells, and screening the targeted single cell clones for blastocyst injection and subsequent embryo transfer. Ultimately, a single survived chimeric mouse with plasmid donor integrated was obtained. This mouse presented a white color hair mixed with black color hair phenotype, indicating that the targeted integration of Tyr gene in Rosa26  locus was correctly expressed.

    Cutting propagation and reintroduction of endangered plant Vincetoxicum hainanense
    XIONG Wujian, SUN Hongmei , LIU Panpan , FENG Lu , GUAN Kailang , LIAO Wenbo , YOU Wenhui
    2020, 59(2):  145-151.  doi:10.13471/j.cnki.acta.snus.2020.02.017
    Asbtract ( 41 )   PDF (1489KB) ( 21 )  
    Related Articles | Metrics
    Vincetoxicum hainanense is a national level Ⅱ rare and endangered plant in China. It has important ecological value. In this study, we selected three main factors, cutting slips, cutting medium and growth hormone that affect the seedling rate of V. hainanense, each factor included 3 levels, the optimum cutting propagation conditions were obtained by orthogonal test consisting of 3 factors and 3 levels. Results showed that: The seedling rate was highest by taking level 3 branches as cutting slips and bamboo grove soil with humus as cutting medium and 0.5 g/L NAA as growth hormone. With regular watering, adventitious roots emerged 7~10 d later and cuttings became seedling 12~15 d later, the seedling rate reached 87.0% after 1 month. We transplanted these seedlings in three natural distribution areas to do a reintroduction test. These areas are primeval forest of V. hainanense, secondary forest of Cinnamomum burmanni and secondary forest of Syzygium jambos. The results showed that reintroduction plants grew very well in primeval forest of V. hainanense, the average survival rate was 967%. After 6 month, the height of plant reached 252.0 cm, base diameter of plant reached 0.78 cm and branch number of the main stem was 3.7, after 9 month, reintroduction plants would blossom and set fruit, the flowering rate and fruit setting rate was 45.5% and 32.7%, it is a feasible method of reintroduction.45.5% and the seed setting rate 32.7%. Together, it can be concluded that it  is a feasible method of ecological restoration.