15 September 2018, Volume 58 Issue 9

    

• Select all
  |

AUTOMATION
• Select
  Survey on light-field displays based on 3D object reconstructions

  TAN Qirui, LU Haiming, LU Zengxiang, JI Yindong

  Journal of Tsinghua University(Science and Technology). 2018, 58(9): 773-780. https://doi.org/10.16511/j.cnki.qhdxxb.2018.22.044

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  Present 3D displays, such as parallax barrier displays, lenticular lens displays, multi-projection displays, and scanning stereoscopic displays, mimic human eyes to obtain 3D images using binocular parallax and persistence of vision. However, locating the 3D images on 2D planes instead of reconstructing the 3D object losses the depth information and causes the vergence-accommodation conflict. 3D objects can be reconstructed using volumetric and holographic displays. However, these require processing of a large amount of data so there are technical limitations due to computing speeds and transmission rates. This paper introduces light-field displays based on 3D object reconstructions. The 3D object is reconstructed using the light field, such as the stacked light field display, shutter light field display, integrated imaging display, or vector light field display. The principles and characteristics of the light field displays are analyzed with the vector light field display having excellent prospects for future development.
  
• Select
  Image recognition and classification by deep belief-convolutional neural networks

  LIU Qiong, LI Zongxian, SUN Fuchun, TIAN Yonghong, ZENG Wei

  Journal of Tsinghua University(Science and Technology). 2018, 58(9): 781-787. https://doi.org/10.16511/j.cnki.qhdxxb.2018.22.034

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  Convolutional neural network (CNN) would easily converge to the local minimum if the network was randomly initialized in image classification tasks. A deep belief network pre-training method was developed by merging unsupervised and supervised methods. Feature sets were extracted from the image patches of zero component analysis (ZCA) whitening and deep belief pre-training to initialize weights of CNNs. Then, convolution features were extracted from the training samples by applying convolution and pooling operations and classified to a specific category through a fully connected network. Finally, the loss value was computed for global optimization. Extensive experimental evaluations on some public datasets show that this method is simple but very effective with the error rate decrease of 0.1% on MNIST and the accuracy increase of 0.56% on Caltech101, which indicates that this method is superior to similar methods.
• Select
  High-speed target selection method for SSVEP based on a dynamic stopping strategy

  WANG Chunhui, JIANG Jing, LI Haiyang, XU Minpeng, YIN Erwei, MING Dong

  Journal of Tsinghua University(Science and Technology). 2018, 58(9): 788-795. https://doi.org/10.16511/j.cnki.qhdxxb.2018.22.038

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  The steady-state visual evoked potential (SSVEP) based brain-computer interfaces (BCIs) are very fast and have huge instruction sets. However, the information transfer rate (ITR) needs to be increased for practical applications. A classification model of the task-relevant component analysis is used on a public data set of 35 subjects to recognize the SSVEP component in the electroencephalography data. A dynamic stopping strategy based on Bayes estimation is used to evaluate the confidence of the classification results. The results show that the dynamic stopping strategy (230 b/min) improves the average ITR by 12.7% compared with the conventional fixed stopping strategy (204 b/min). Thus, this result shows how SSVEP-BCI can be further improved by the Bayes-based dynamic stopping strategy.
  
• Select
  E-commerce market operation monitoring service platform

  CHAI Yueting, SUN Baowen, ZHANG Anting, LIU Hang, YU Chao

  Journal of Tsinghua University(Science and Technology). 2018, 58(9): 796-801. https://doi.org/10.16511/j.cnki.qhdxxb.2018.25.035

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  The government needs to identify e-commerce developments and trends to develop effective regulations. This study analyzed key features of emerging e-commerce marketplaces, designed a monitoring system for e-commerce marketplaces, and developed the operations monitoring indicators and the platform. The platform not only provides government agencies a decision-making service to understand the real-time e-commerce status and trends, but also provides a foundation for a complete monitoring system for all future economic activities as the e-commerce economy continues to be combined with the traditional economy.
• Select
  Design and implement of a public service platform for e-commerce credible transaction guarantee

  CHAI Yueting, YU Xiao, LIU Zhenming

  Journal of Tsinghua University(Science and Technology). 2018, 58(9): 802-807,820. https://doi.org/10.16511/j.cnki.qhdxxb.2018.22.039

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  The information and behavior of e-commerce market transaction subjects, objects, processes and results need to be standardized to make the systems trustworthy. In addition, the systems need to eliminate illegal, fake and shoddy information and behavior in the e-commerce market to improve the e-commerce market credibility. This paper describes the basic principles for guaranteeing that the e-commerce transaction subjects, objects, processes and results are credible. The overall structure and function of an e-commerce credible transaction guarantee public service platform are described with the system then implemented and verified. This e-commerce model provides market norms, service and supervision for standardized management, accurate inspection, social co-governance and collaborative law enforcement.
• Select
  Effects of social interaction factors on a consumers' eWOM motivation and intention in e-commerce websites

  ZHANG Lin, YAN Qiang, LIU Yi

  Journal of Tsinghua University(Science and Technology). 2018, 58(9): 808-813. https://doi.org/10.16511/j.cnki.qhdxxb.2018.25.036

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  Social interaction factors affect consumer eWOM behavior and can promote product sales and improve the services of e-commerce websites. This study analyzed 6 kinds of eWOM motivating factors including entertainment, self-enhancement, communication and sharing, supporting or punishing merchants, improving service and economic incentives. A questionnaire was designed to study the relationships among these social interaction factors, the eWOM motivation and the eWOM intention of e-commerce websites. The results show that social interaction factors positively affect eWOM motivation and positively affect eWOM intention. This study provides an empirical basis for describing and predicting consumer online behavior with suggestions for the design and operation of e-commerce websites and review systems.
ENERGY AND POWER ENGINEERING
• Select
  Effect of SO₂ and CO₂ on NO removal efficiency by the hexamminecobalt(II) complex in a bubble column

  LIU Zhen, LI Qinghai, ZHU Qunyi, TAN Zhongchao, ZHANG Yanguo

  Journal of Tsinghua University(Science and Technology). 2018, 58(9): 814-820. https://doi.org/10.16511/j.cnki.qhdxxb.2018.22.032

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  The effects of SO₂ and CO₂ in the flue gas and their reaction products with ammonia, SO₃^2- and CO₃^2-, on the NO removal efficiency of the hexamminecobalt(Ⅱ) complex were studied experimentally using a bubble column reactor with the hexamminecobalt(Ⅱ) complex absorbing NO. The effects of SO₂ and CO₂ on the NO removal efficiency were studied using various SO₂ and CO₂ concentrations in the flue gas with the effects of SO₃^2- and CO₃^2- on the NO removal efficiency studied with various concentrations of SO₃^2- and CO₃^2- into the hexamminecobalt(Ⅱ) solution. The experimental results show that SO₂ and CO₂ in the flue gas reduce the NO removal efficiency with the NO removal efficiency reduced 46% for a CO₂ mass concentration of 100 g/m³ and an SO₂ mass concentration of 2 646 mg/m³. The SO₃^2- and CO₃^2- formed in the solution also reduce the NO removal efficiency. The NO removal efficiency is reduced by 72% for an SO₃^2- molarity of 0.03 mol/L and by 41% for a CO₃^2- molarity of 0.50 mol/L.
NUCLEAR AND NEW ENERGYT ECHNOLOGY
• Select
  Thermodynamic analysis of closed Brayton cycles for space reactor power system

  YANG Xie, SHI Lei

  Journal of Tsinghua University(Science and Technology). 2018, 58(9): 821-826. https://doi.org/10.16511/j.cnki.qhdxxb.2018.22.035

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  Helium-xenon mixtures can be used as the cycle working fluid and reactor coolant for space nuclear reactor power (SNRP) systems using a closed Brayton cycle. The cycle designers must know how the components of the helium-xenon mixture affect the net system efficiency to choose the best working fluid. A thermodynamic model is developed with Fortran 95 for the SNRP Brayton cycle to analyze the net system efficiency for various molecular mass mixtures in terms of the adiabatic coefficient, regenerator effectiveness and normalized pressure loss coefficient. The results show that the adiabatic coefficient has little effect on the net system efficiency, while the net system efficiency increases with increasing regenerator effectiveness and decreases with increasing normalized pressure loss coefficient. The compressor and turbine in a space Brayton cycle are smaller with higher molecular mass helium-xenon mixtures, so He-8.6%Xe is chosen as the cycle working fluid with the maximum regenerator effectiveness. For cold and hot sink temperatures of 403 K and 1 300 K, the net system efficiency is 29.18%.
AEROSPACE
• Select
  IQ-separated baseband structure for high-speedsatellite digital communication

  PEI Yukui, HAO Haoran, SU Li

  Journal of Tsinghua University(Science and Technology). 2018, 58(9): 827-832. https://doi.org/10.16511/j.cnki.qhdxxb.2018.22.031

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  Increased satellite digital communication rates require larger channel bandwidths which require analog-digital converters (ADCs) with higher sampling rates in the receiver. The in-phase/quadrature (IQ) data have to be sampled separately for ultra-high sampling rates, which requires IQ-separated processing. A baseband structure developed here for IQ-separated processing uses redesigned pilot arrays, an adjusted processing order and modified algorithms that gives separate frequency error estimates and timing synchronization on IQ data. Simulations show that this structure is fast with a bit-error-rate only 2 dB higher than the traditional structure.
• Select
  Occurrence probability of co-frequency interference and system availability of non-geostationary satellite system in global dynamic scene

  JIN Jin, LI Yaqiang, ZHANG Chen, KUANG Linling, YAN Jian

  Journal of Tsinghua University(Science and Technology). 2018, 58(9): 833-840. https://doi.org/10.16511/j.cnki.qhdxxb.2018.25.038

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  With the increase in the number of non-geostationary orbit (NGSO) satellite systems, the interference between NGSO satellite systems using the same frequency has become increasingly prominent. Because of the large number and time-vary relative motion of NGSO systems, the traditional interference analysis methods and evaluation index, which are aiming for geostationary orbit (GSO) satellite systems, are no longer suitable for NGSO scene. Based on the relevant rules and recommendations of the International Telecommunication Union, this paper establishes a mathematical model of the interference analysis for non-geostationary satellite systems, and proposes a link angle probabilistic analysis method for satellite constellation interference analysis. Aiming at the complicated features such as the large number of satellites consisting of NGSO constellations and the changing temporal and spatial relationship, the scheme of NGSO constellation global interference analysis is given. The probability calculation method of harmful interference between satellite constellations and the index of satellite constellation usability are proposed. On the basis of the actual satellite network data, taking OneWeb system and O3b system as an example, the range of the link angle of the interference protection between the satellite systems is calculated. The angle of the satellite link, the interference state and the probability distribution of the availability ratio on a global scale are given, provides a means for NGSO constellation interference analysis.
CONSTRUCTION MANAGEMENT
• Select
  Reinsurance auction for surety bonds based on multi-attribute reverse auctions

  XUE Yanguang, DENG Xiaomei, SU Guiliang

  Journal of Tsinghua University(Science and Technology). 2018, 58(9): 841-848. https://doi.org/10.16511/j.cnki.qhdxxb.2018.22.045

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  Surety companies always face the problem that the guaranteed amount per bond or the guaranteed balance often exceeds its limit. Surety companies then need surety reinsurance but such policies still lack operational guidance on the use of surety reinsurance in China. This paper presents the surety reinsurance system as a competitive surety reinsurance platform and introduces an auction mechanism for the reinsurance operations to increase the reinsurance efficiency and competitiveness and promote market-oriented reinsurance premiums. Then, a surety reinsurance game model is developed based on a multi-attribute reverse auction with a Bayesian-Nash equilibrium strategy. Its three bidding attributes for the candidate reinsurer include the total reinsurance expense, the underwriting risk and the credit rating from the credit reporting system. A Monte Carlo simulation is then used to simulate the model and analyze the trends of the total score and the sub-attribute scores of the winning candidate for various numbers of candidates. The simulation results agree well with real data.
• Select
  Impacts of land supply planning on local government land supply behaviour: An analysis based on prospect theory

  WU Shuping, YANG Zan

  Journal of Tsinghua University(Science and Technology). 2018, 58(9): 849-857. https://doi.org/10.16511/j.cnki.qhdxxb.2018.26.034

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  The land supply significantly impact economic and social development. China supplies state-owned land under an administrative plan whose effectiveness determines the land resource allocation efficiency. Prospect theory is used here to investigate a local government's residential land supply behavior for one representative administrative plan from 2010 to 2013 based on panel data from provincial and prefecture level cities in China. The results illustrate the reference dependence and diminishing sensitivity of a local government's residential land supply behavior and indicate that a residential land supply administrative plan will not be efficient if the objective is set too high.
ENGINEERING PHYSICS
• Select
  Scenario model for emergency decision-making

  ZHU Wei, LIU Cheng, LIU Yi

  Journal of Tsinghua University(Science and Technology). 2018, 58(9): 858-864. https://doi.org/10.16511/j.cnki.qhdxxb.2018.22.037

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  This paper presents a disaster scenario model that gives a structured, quantitative picture of the disaster status to improve situation awareness during a disaster. A multi-dimensional scenario model is developed for emergency decision-making. The scenario model includes object representations, disaster status vectors, the scenario context and a scenario solution. The object representations and disaster status vectors are the core of the scenario model as qualitive and semi-quantitative representations of the disaster status. The scenario context and scenario solution are structural descriptions of the disaster environment, response resources and emergency response actions. Mathematical definitions and formulations of the scenario model are given with an example to demonstrate the model effectiveness. This model defines the scenario with the disaster status represented by a structured, quantitative model. The model can be used to evaluate the emergency severity for decision-making.