• ISSN 0258-2724
  • CN 51-1277/U
  • EI Compendex
  • Scopus
  • Indexed by Core Journals of China, Chinese S&T Journal Citation Reports
  • Chinese S&T Journal Citation Reports
  • Chinese Science Citation Database

2005 Vol. 18, No. 3

Display Method:
mici
Research Advances in Automatic FacialExpression Recognition
ZHANG Jia-shu, CHENHui, LIDe-fang, LUO Xiao-bin, XIAXiao-dong
2005, 18(3): 285-293.
Abstract:
As a challenging interdiscipline ofbiologic feature recognition and affection calculation, the technique ofautomatic recognition of facial expression (facial expression automatic recognition system, FEARS) develops quickly driven by demands of various applications. Nevertheless, fully automatic facial expression recognition systemswith acceptable robustnesshave notyetcome forth due to the great difficulties. The three key procedures for automatic recognition of facial expression are facial image detection, location and extraction of facial features and emotion classification, and greatadvances have been achieved on the three fields. The remaining problems include improving the robustness of facial recognition algorithms, precise and pertinence of identification of facial features, separation of rigid facial actions and recognition of three-dimensional facial expressions. To build the data bases of facial expressions and application research are also importan.t Indexes, such as applicability, comparability and appropriative and real-time performances, are proposed for evaluating automatic recognition systems of facial expressions.
Technical Schemes forM etro Station Construction Based on ShieldM ethod
HE Chuan, DING Jian-long, LIWei
2005, 18(3): 293-297.
Abstract:
The construction methods of overseas metro stations driven with the shield method were introduced. Based on the fact thata single-line shield tunnelhas a diameterofabout6. 0m in China, four construction schemes of ametro station driven from a sectional shield tunnelwith the expanding excavationwere proposed. According to the main technical standards and construction dimensions of Linhexilu station in the Guangzhou No. 3 metro line, the structures and construction schemes of this stationwere designed. Finally, the main related technical problems to be resolved were discussed, including construction steps, design method of station structure, design of special segmen,t environment control and project cost.
3D Elasto-plastic Numerical Simulation of Entire Construction Process for H ighway Tunnel
SHE Jian, HE Chuan
2005, 18(3): 298-302.
Abstract:
In order to obtain the regularities of stress, strain and displacement of tunnel structures during excavation, the northern end ofBijiashan tunnel in the 2nd stageYu-Qian projectwasmodeled and its construction process was numerically simulated with 3D elasto-plastic finite elementmethod (FEM). Comparison between simulated and in-situ tested results shows that the range with a great influence of tunnel excavating on surrounding rock is5m around an excavation face, excavation of the upperbench has a greater effect than that of the lower bench, and one-third of total settlement and horizontal displacement of surrounding rock has generated before excavating. In addition, the plastic zone in surrounding rock ranges 1. 5 m in the front of an excavating face, and rock bolts are mainly influenced by the excavation of the upperbench in the current construction step.
CFD Simulation ofAir Distribution of Subway Side-Platform
YUAN Feng-dong, YOU Shi-jun, YANGXiang-jin
2005, 18(3): 303-307.
Abstract:
To provide a reference for the airdistribution design ofa subway side-platform and appraise the blue print of an air distribution design, a three-dimensionalmodel for typical side-platformswas constructed to got an appropriate supply air temperature difference to optimize the design of an air conditioning system of a subway side-platform. The physical model is the standardk-εturbulence mode.l Based on actual heat load and wet load the boundary conditions for the numerical simulation were determined. The air flow ofa subway side-platform was simulatedwith CFD (computational fluid dynamics). The result shows that the simulated temperature field is the same as the tested temperature field. This indicates thatCFD can simulate and forecast the temperature field ofa subway side-platform accuratelywith the standardk-εmode,l and then provide a reference for the design of the air- conditioning system of subway side-platforms.
M odelingM ethod for Large Strip DigitalTerrainM odel
YISi-rong, L Xi-kui, ZHANG Jia-ling
2005, 18(3): 308-312.
Abstract:
From the characteristics ofmass terrain data in a railway location system, a practical fast algorithm for strip digital terrain model (S-DTM) generation was proposed based on the constraint DelaunayTIN (triangulated irregularnetwork). Thismethod takesmany of the characteristics of large strip terrain areas into consideration, such as obtaining the boundary ofsub-areas, finding outpoints at different sides, shortestdistance, characteristic lines, and special areas, etc. The decomposition and structure ofquad-tree are used to large triangulated netmodels constructed to increase the efficiency of the generation and application of S-DTMs. This method has been used in a virtual environmental railway location system. The result of application in several practical railway lines shows the validity and reliability of the proposedmethod to establishing continuous strip digital terrainmodels.
GPSM onitoring ofX ianshuihe Fault and Blocks on Its Both Sides
TANG Wen-qing, LIU Yu-ping, CHEN Zhi-liang, HUANG Ding-fa
2005, 18(3): 313-317.
Abstract:
Based on the GPS measurement of the Xianshuihe fault and blocks on its both sides, Chuan-Dian block and Chuan-Qing block, a velocity field on the Eurasia reference frame was obtained. The movement velocities ofGPS stations in the region were estimated using least-squares fitting under the hypotheses of rigidmass. The fitting result shows that themovementvelocities of the Chuan-Dian and Chuan-Qing blocks are 19. 5±3. 4 and 13. 2±3. 4 mm/a respectively, and the Chuan-Qing block moves at a speed of 8. 3±3. 4 mm/a relative to the Chuan-Qing block. The movementvelocity of the Xianshuihe fault is 7. 9 to 8. 8 mm/a( it is 8. 8±3. 5 mm/a in Garzê- Qianning segment and 7. 9±3. 3 mm/a in Qianning-Kangdin segment). The above results show that the region covered the Xianshuihe faul,t the Chuan-Dian and Chuan-Qing blocks is an important tectonic activity zone.
M easurementM odel ofDigitalM aps Based on Geographic Coordinates
YANG Yong-chong, GUO Da-zhi
2005, 18(3): 318-321.
Abstract:
To construct a digitalmap based on geographic coordinates is a bettermethod to bring digitalmaps in different coordinate systems or projection zones into one system or joint them. In the measurementof digitalmaps based on geographic coordinates, the deviation between the reference ellipsoid surface and the earth s surface results in the discrepance between the calculating value on the reference ellipsoid surface and the real surveying value on the earth s surface. Because a projection plane ismore close to the earth s surface, the calculating value on the projection plane ismore close the real surveying value on the earth s surface. Therefore, themeasurementmodelofdigitalmapsmay be setup by establishing a compensation coordinate system. In thismode,l the compensation projection plane is the surface ofaverage elevation ofcalculated objects, and the centralmeridian ofprojection is the average longitude line of calculated objects, so the calculating result is most close to the real surveying value. This model (method) is more suitable for the regions with a small area and a requirementofhighmeasuring accuracy.
Influence ofRainfall on Stability ofCarrying Soil ofAfforested Slope
WUXi-yong, LIANG Yi, LIShu-ding
2005, 18(3): 322-324.
Abstract:
Based on themodel of carrying soil failure under the condition of rainfal,l the influence of rainfallon the stability ofcarrying soilduring the afforestation ofa slopewas investigated. The research indicates that the stability of carrying soil is closely related to the character of rainfal,l the gradientof an afforested slope, the characteristics of slope surface, the feature of soil and the thickness of soi.l When the gradientof a slope to be afforested is large, a thin soil layerwill be necessary in order to maintain the stability of carrying soi.l
Upper Lim it ofAngle between Strikes of Slope and Strata for Dip Slope
BAIYun-feng, ZHOUDe-pei, FENG Jun
2005, 18(3): 326-329.
Abstract:
Similar simulation testsweremade to determine the upper limitofangle between the strikes of slope and strata for a dip slope. Furthermore, the developmentcharacteristics ofbedding-slips along the Chongqing-Huaihua railway from Fuling City to Longtan Town section were investigated to prove and complement the experimental result in lab. The research result shows that the stability of a dip slope enhanceswith the increase of the angle between the strikes ofslope and strata, and bedding-slips occur seldom when the angle is over 30°. Therefore, from the experimental and statistical results the upper limitof the angle for a dip slope is 30°.
Role-Based Access Control ofCollaborative Design System s
LIBai-lin, LUMao-hua
2005, 18(3): 330-333.
Abstract:
The roles were analyzed in terms of their characteristics in the access control of collaborative design systems, and their granularity were determined. The technique of role-based access control (RBAC) and session function provided by JSP (Java serverpages) were used to realize an access control system. The access control system was applied in a prototype system of collaborative design. The trial results show that the proposed system satisfies the requirementofcollaborative design system.
Performances ofUltra-fine TiC0. 7N0. 3CermetCutting Tools
GONG Bang-ming, YU Jun-hui
2005, 18(3): 334-337.
Abstract:
Cutting performances of ultra-fine TiC0. 7N0. 3cermet cutting tools were investigated and compared with those of ordinary cermet and carbide cutting tools. The exponential formulas of the cutting forces were obtained based on the experimental data. The lives of the ultra-fine TiC0. 7N0. 3 cermet cutting tools under study are 2. 5 and 3 times as long as those of ordinary cermet and carbide cutting tools, the cutting forces decrease by 9% to 10%, and the machined roughness is as low as 0. 7μm, obviously lower than those of ordinary cermet cutting tools (1. 3μm) and carbide cutting tools (1. 2μm).
G lobalExponential Stability ofHopfield NeuralNetworks
TANXiao-hui, ZHANG Ji-ye, YANG Yi-ren
2005, 18(3): 338-342.
Abstract:
The existence, uniqueness of the equilibrium point of a class ofHopfield neural networks were studied by introducing nonlinearmeasures. The activation function is nonlinear and notnecessary to be continuous, monotone and bounded. Some sufficient conditions and convergence estimation for global exponential stability ofneuralnetworkswere obtained.
G lobalExponential Stability of a C lass of Interconnected Large-Scale System swith Time Delays
RENDian-bo, ZHANG Ji-ye
2005, 18(3): 343-346.
Abstract:
The global exponential stability of a class of linear interconnected large-scale systemswith time delayswas analyzed based onM-matrix theory and by constructing a vectorLyapunov function. A criterion was obtained for global exponential stability of the systems by analyzing the stability of differential inequalitieswith time delays under the assumption that the time delays are bounded and continuous. A large-scale system is global exponential stable if the testmatrix is anM-matrix, where the testmatrix is constructed by employing the coefficientmatricesof the system and the solutionsof the Lyapunov equationswhich are interconnectedwith the system. Since it is independentof the delays and simplifies the calculation, the criterion is easy for application.
Experimental Investigation onM echanicalProperties of Nano-Al2O3Coating on ResinM atrix Composites
XULi, JIANGXiao-yu, QIAN Lin-mao
2005, 18(3): 347-350.
Abstract:
The mechanical properties of nano-Al2O3coating on Al2O3short fiber reinforced epoxy composites were analyzed with nano-indentation and nano-scratch tests to study the relationship between the properties and coatings and coating procedures. The procedures to coat the coatingsduring and after the solidification of the epoxywere used. The former and latter proceduresmake the elastic modulus of the coatings increase from 1. 0 GPa for thematrix to 3. 3~3. 9 GPa and 2. 1~4. 0 GPa, respectively, and the hardness of the coatings raise from 200MPa for thematrix to 700~1 300MPa and 400~600MPa, respectively.
ThermalResidual Stresses and Their Effects on Properties ofδ-Al2O3 Short Fiber Reinforced AlAlloy Composites
ZHANG Juan, GAO Qing, KANG Guo-zheng
2005, 18(3): 351-354.
Abstract:
A double fibermodel was proposed. The distribution of thermal residual stresses inδ- Al2O3 f/Al alloy composites and the effects of thermal residual stresses on the properties of the compositeswere calculatedwith a double fibermodel and elasto-plastic finite elementmethod. Tensile stress-strain curves of threeδ-Al2O3/Al alloy composites were obtained with the effect of thermal residual stresses taken into account and compared with the experimental results. The effects between neighboring fibers result in the changes in magnitude of thermal residual stresses, but have no influence on their space distribution. Reasonable consistencywas achieved between the simulated and experimental stress-strain curves forδ-Al2O3/Al-5. 5 Mg andδ-Al2O3/Al-5. 5 Zn composites, but therewas significant error in the simulated curves comparedwith the experimental results forδ-Al2O3/ Al-12 Si composite.
VerticalRandom Response Analysis ofLocomotives Based on AugmentedM atrix
YANG Guang-wu, XIAO Shou-ne, JINDing-chang
2005, 18(3): 355-340.
Abstract:
Augmentedmass, damping, stiffnessmatrix ofsystem was setup to derive verticaldynamic differential equations for locomotive components and obtain output responses of suspensions. To avoid cross-spectrum density of each inputwhen calculating spectrum density of system outpu,t multi-input excitation in six wheelsetwere converted to single-input excitations by making use of the time-delay between the input excitations inwheelsets, making the system a linear system with single-input-multi- outpu.t Frequency response functions of the converted system were obtained according to superposition principle of linear systems and spectrum analysis of frequency responsematrix. The relations between spectrum densities ofsystem outputresponses and frequencywere obtained by assuming input functions to be ergodic and stable stochastic processes.
Time Restraint-BasedM odel and Algorithm for Railway EmptyW agon Distribution
LIZong-ping, XIAJian-feng
2005, 18(3): 361-365.
Abstract:
Traditionalmathematical optimalmodels for empty wagon distribution do not consider the time efficiency of emptywagon distribution and efficientusage ofemptywagons by loading them where they are or in the neighboring stations. To overcome the shortcomings of the traditionalmodels, a new model and its algorithm were proposed. In the proposedmode,l time restraintwas taken into accoun,t and the distribution ofboth loaded and emptywagons is calculated bymaking good use of the available emptywagons in the stations where they are to be delivered to decrease the delivering distances of emptywagons. The feasibility of the proposedmodel and algorithm was verifiedwith an example.
Evaluation and Enhancement ofTransportCapacity of Railway System in BAOSTEEL
MAGuo-zhong, LIUQing-wei, WANG Ci-guang
2005, 18(3): 366-370.
Abstract:
The transport capacity of the railway system in BAOSTEEL was evaluated based on a thorough investigation on the carrying capacities of the sub-systems, such as locomotive, arrival- departure tracks, and loading-unloading of products and coke. The investigation shows that the subsystems of locomotive and arrival-departure tracks form the bottleneck of the railway system, the annual transportation capacities ofwhich are 2. 91million tand 2. 96million ,t respectively. Adding a locomotive and a arrival-departure track will raise the annual carrying capacities of the two limiting subsystems to4. 39million tand 3. 56million ,t respectively, tomeetthe demandsofcarrying capacity by 2009.
EconometricM odel for Transportation System of Sichuan Province
FANG Gui-jin, PENG Yue
2005, 18(3): 371-374.
Abstract:
An econometric model consisting of 12 explicit formulas for analyzing the transportation system ofSichuan province, China, was derived according to the principles ofsystems engineering and system dynamics. Themodelwas tested in terms ofeconomic significance, statistics, econometrics and stability, and the results verify that the it conformswith the statistical economics tes.t The statistical data of the transportation system ofSichuan Province from 1992 to2001were taken as the inputs for the mode.l The error rates during the dynamic simulation are less than 20%, and the error rate is4. 17% for thewholemode,l indicating reasonable precise and stability of themode.l
Evaluation of Sustainable Development in CoalEnterprises Based on BP NeuralNetwork
XU Jun, DUWen, ZENG Qi
2005, 18(3): 375-379.
Abstract:
Based on analyzing the existingmethods for evaluation of sustainable development of coal enterprises, an evaluation-index system was established. A neuralnetworkmodelwithmulti-hierarchic structurewas built based on error back-propagation (BP) algorithm. The conjunction weights of the neural network are modified layer by layer from output layer to input layer in the process of neural network training to reduce the errors between the anticipated and actual outputs. A case study was conducted and itverified the validity and accuracy of the proposedmethod.
Analysis on RegionalVenture CapitalEnvironmentBased on AHP Fuzzy Comprehensive Evaluation
JIN Jing-yu, LIUChao-ming, HAN Bin
2005, 18(3): 379-384.
Abstract:
AHP ( analytic hierarchy progress) and fuzzy evaluation were combined to get a comprehensive evaluation method for evaluating venture capital environments of big cities in China, where there have been fastdevelopments in venture capita.l A big city inwestChinawas taken as an example. The evaluation index system in the example consists of three layers: an objective layer, a rule layerwith 8 factors, and an index layerwith 36 indexes. The evaluation resultshows thatexplicit evaluation results can be obtainedwith the proposedmethod and scattering of results due tomultilayer and complexity ofventure factors can be avoided.
Game Analysis of Stock PriceM anipulation Through Takeover Bids
QIUPei-guang, ZHANG Yong-peng
2005, 18(3): 385-389.
Abstract:
The influences of an increase in noise trading before and after bidding and the level of takeover activities on the possibility ofmarketmanipulation were analyzed with an incomplete game model to explore ifdropping bid should be allowed after a buyer has bid. The conclusion is that the unique equilibrium leads to a non-profitable manipulation, and social surplus is increased by the possibility of dropping bids, if the level of takeover activity is high; on the contrary, the unique equilibrium results in a profitable manipulation, and social surplus is decreased by the possibility of dropping bids, if the level of takeover activity is low.
Application of Jar-M etric Principle for Solving Lexicographic OrderMultiobject andMultistage Decision Problem s
WANGMing-hui
2005, 18(3): 390-393.
Abstract:
A lexicographic ordermultiobjectandmultistage decisionmodelwas derived. The principle and algorithm of Jar-metricwere applied to solving the problem. In a Jar-metric graph representing a lexicographic ordermultiobject andmultistage decision problem, a Jar valve from a starting node to a destination node is equal to a monoid sum from the starting node to the destination node, and the monoid sum is independentof the nodes before the starting one and those after the destination one. In the calculation, conjunction is taken as Jar summation, and ordinary summation is regarded asmonoid produc.t In this way, the recursive calculation for solving lexicographic order multiobject and multistage decision problem is transformed to amultistage Jarmatrix computation.
Algorithm for FaultLocation ofElectric Traction Network with Series Compensators
WANG Ren, LIN Guo-song, QIANQing-quan
2005, 18(3): 394-397.
Abstract:
The relationship between impedance and distance of power traction network with series compensators is notmonotone in the electric traction network of electrified railways. The conventional method of fault location fails because therewillbe two ormore faultpoints corresponding to an actual fault. A newmethodwasproposed based on analysison the faultmodelofelectric traction networkwith series compensators. ATP (alternative transients program) simulation programmed withMATLAB on different faultphase angles and transient resistances indicates that the method is able to identify the position relative to the series compensators. The algorithm was presented to help users to apply the proposedmethod by looking up a table containing the criteria.
M icro-Computerization ofFrequency-Shifting Automatic Block Transm itting Equipment
XIANG Jun
2005, 18(3): 398-401.
Abstract:
To overcome the shortcomings of traditional equipment of frequency-shifting automatic block, a three-display frequency-shifting automatic block transmitting device was designed. It uses a single-chip microcomputer as the main componen,t an external timer/counter and some simple auxiliary circuits to produce frequency-shifting and low frequency modulation signals. This design provides the equipmentwith high reliability and flexibility. It can also adapt itself to the transmitting equipmentwith 4-, 8-even 18-displayswith a slight extension ofhardware and simplemodification of software.
FastAtom Construction Algorithm for SignalDecomposition in Over-Complete Dictionary
HUAZe-xi, YIN Zhong-kel HUANGXiong-hua
2005, 18(3): 402-405.
Abstract:
It is one of themain problems in signal decomposition in over-complete dictionary that the atom construction process is very slow. To solve this problem, a new fast algorithm was proposed. In the algorithm, all atoms are divided into two categories: small and large atoms, according to their scales. Because the energy of a small atom concentrates in a small region of the whole atom, it is constructed within the region ofenergy concentration, and thewhole atom is represented by the locally constructed atom. A large atom is constructed by interpolation after a corresponding small atom has been constructed. Experimental results show tha,t when the length of the signal is 256, the proposed algorithm is 4. 7 times faster than traditionalmethodswith the same signal quality.
Knowledge Base System for DataM ining
ZHUHeng-min, JIXiao-li, WANGNing-sheng
2005, 18(3): 406-412.
Abstract:
Six types of domain knowledge for data mining were summarized. A method of two-level knowledge organization, that is a combination of index knowledge and domain knowledge, was proposed. The former, as the knowledge catalog, codes the latter according to the characteristics of knowledge. Domain knowledge base consists of rule, function, hierarchical knowledge and basic information sub-bases. Syntax check for allkinds ofdomain knowledgewas discussed. Itwas pointed out that conflict forms of domain rule included conclusion conflic,t confidence conflict and inclusion conflict, and redundancy forms ofdomain rule included equivalence redundancy, inclusion redundancy and condition redundancy. Check algorithmswere presented subsequently. A knowledge base system, namedKB4DM (knowlege base for datamining), was implemented based on the research.
Comprehensive EvaluationM odel for Difficulty Coefficient ofRoad Adm inistration
JIANG Gao, SONGXue-jun, CHENDe-xin
2005, 18(3): 412-416.
Abstract:
The factors affecting the difficulty coefficient of road administration were analyzed, and a new model was proposed. In the mode,l the index system is established based on AHP (analytic hierarchy progress), and the road administration information is standardized based on fuzzy logic theory. Theweights from experts are tested with group consistency algorithm to obtain the normalized weights. Therefore, the difficulty coefficient of road administration is determined by calculating the group utility function based on the normalizedweights, utility function ofevery expertand the indexes.
A NovelP2P Resource Sharing IncentiveM echanism
XIAO Bo, JINWei, HOUMeng-shu
2005, 18(3): 417-421.
Abstract:
To resolve the“free rider”and the“tragedy of the commons”problem in peer-to-peer storage systems. In this paper, we propose incentivemechanism in aP2P storage sharing system based on the genetic algorithms. In this mechanism, the provider allocates the sharing space based the genetic algorithms, and maximum its contribution by choosing the optimum policy. The more contribution and themore sharing storage which its can use. Analyses and simulations show that the genetic algorithms-based P2P incentivemechanism is fairness and effective, and itcan discard the free- rider from P2P system effectively and encourage the node sharingmore storage so that itcan earnmore reward.
Sequence Design ofZero Correlation Zone Based on Interleaved Technique
WANG Long-ye, TANGXiao-hu
2005, 18(3): 422-425.
Abstract:
To lower the complexity of the existing methods, a new sequence design method of zero correlation zone (ZCZ) based on interleaved technique was proposed. For a given ZCZ sequence set where the length isL, the size of the sequence isM, and the zero zone isz, a new sequence set is obtained bymtimes ofrecursive constructionsusing interleaved technique. The length andwidth of the new sequence set is 2mtimes as large as those of the original one, and the size of the sequences remains the same asM. The new ZCZ sequences are easily implemented in terms of hardware and software, and is suitable for the approximately synchronous CDMA (AS-CDMA) communication system.
M atched Filtering ofLFMCW Radar Signals
GUO Lei, YANG Zhong-hai
2005, 18(3): 426-429.
Abstract:
The outputofmatched filters for linear frequencymodulation continuouswave (LFMCW) signalswas analyzedwith period ambiguity function. Frequencyweighting is effective to depress range side lobes and improve multi-target resolution, but itwidensmain range lobes in the same time and leads to loss in signal to noise ratio (SNR) of the output ofmatched filters forLFMCW signals. The loss ofSNR is 3. 52 dB, by takingHanningweight as an example.
Focal Shift in Elegant sin-Gaussian Beam s Using Second-OrderM omentM ethod
HUANG Yu-lin, WANG Li, KONG Rui-xia, WANGXi-qing
2005, 18(3): 430-434.
Abstract:
The focal shiftof elegant sin-Gaussian beams passing through a lenswas studied using the Collin s formula. Itwas found that the axial intensity intensity of a focused elegant sin-Gaussian beam was nul.l The second-ordermomentmethod was used to derive the analytical formulas for both beam waist and the position of thewaistof the focused elegantsin-Gaussian beam. The focal shifts ofelegant sin-Gaussian beams can be obtained using the derived formulas. The analytical formulas for the relative focus shifts of focused elegantsin-Gaussian beamswere presented as examples. Numerical calculations demonstrate that focal shifts of focused elegant sin-Gaussian beams depend on optical system parameters, beam parameters and Fressnel number. Proper choice of the distance from the incidence plane to the lens leads to disappearance of focal shift.