西南交通大学学报

Track Irregularity Analysis Based on Wavelet Transform

ZHANGJie, CHENChun-jun, LINJian-hui

2004, 17(5): 469-471.

[Abstract](2241)

Abstract:
To eliminate the disturbances from the speed-instability of the measurement vehicle on the collected track irregularity spectrum, Haar wavelet transform was used to analyze the instantaneous frequencies of the track irregularity signals. The relationship between the instantaneous frequencies of the track irregularity signals and the Haar wavelet coefficients was established to calculate the instantaneous frequencies. The calculated instantaneous frequencies are used for interpolation to get the re-sampled signals, which are then Fourier-transformed to get disturbance-free frequency spectrum.

Bound Searching Algorithm for Shortest Path in a Network

LI Yin-zhen, GUO Yao-huang

2004, 17(5): 561-564.

[Abstract](1719)

Abstract:
Many irrelevant vertexes must be searched in order to find the shortest path between two vertexes in a large networkwith Dijkstra algorithm, resulting in low efficiency. To improve the searching efficiency, a new searching algorithm, named bidirectional bound searching algorithm, was proposed. With this method, a shortest path from the source vertex to the terminal vertex via any vertex is determined first by bidirectional searching. This path is closed to or even the same as the shortest path to be found. The length ofthis path is then taken as a bound, and any vertex with larger label value than the bound in later calculations will not be considered. Itwas proved that the calculating efficiency of the newalgorithmwas as twice as that of the traditional Dijkstra algorithm

Strong Kernel Graphic Algorithm for Searching All Essential Circuits of Simple Directed Graph

ZHAO Yu-hua, LIKe-bai, RENWei-min

2004, 17(5): 565-568.

[Abstract](1541)

Abstract:
The problem of searching all feedback loops of a system dynamics model is equal to calculating all essential circuits of a corresponding simple directed graph. The keyof the problem is time complexity. A newconcept named strong kernel was defined, and the algorithm based on strong kernel for searching all essential circuits of a simple directed graph was proposed. An illustrative example was presented and the complexity of the algorithmwas analyzed. In terms of time complexity, the proposed algorithm is superior to that of the determinant algorithm based on the concept of kernel of directed graphs.

Tail-Number-Assignment Problem of Airliners Based on Fleet Balance Application

SUNHong, WENJun, XUJi

2004, 17(5): 569-572.

[Abstract](1404)

Abstract:
To solve airliner tail-number-assignment (TNA) problem with the requirement of balanced application of airplanes, a simulated annealing algorithmwas designed. According to the character of single hub and spoke network widely used in Chinese airliners, a new concept called flight pairing was defined. When sufficient aircrafts are available, constraint on application of each aircraft is ignored. By constructing a flight pairnetworkmodel describingthe connectingrelationship of flight pairs, the originalTNAproblem is transformed to optimizing flight pair schemes with minimized target function. A simulated annealing algorithmwas presented. The computing complexity of this algorithm isO(M(m3+mn)).

Road Network Capacity Reliability Based on Link Travel Time Reliability

LIUHai-xu, PUYun

2004, 17(5): 573-576.

[Abstract](1417)

Abstract:
To decrease the amount of calculation, a bi-level programming model based on link travel time reliabilitywas set up for road network capacity reliability through analyzing the concept of road network capacity reliability. Based on the character of route choice shown in the model,i.e.user equilibrium, an algorithm based on the sensitivity analysiswas used to solve the model. Finally, a simple numerical example was given. The numerical result shows that the model can be used to estimate the performance of road networks under uncertain environment and provide a way to improve the performance of road networks.

Optimal Approval Model for Wagon Requisition in Railway Bureaus

WANG Wei, YAN Yu-son, WANG Yong, ZHAONan

2004, 17(5): 581-584.

[Abstract](1547)

Abstract:
An optimal approval model for wagon requisition in railway bureaus and the corresponding network were proposed. In this way, the problem of optimizing approval for wagon requisition was transformed into solving minimum cost and maximum flow in a network. A numerical algorithmwas derived to solve the model. Several solutions to some actual problems were put forward. The software based on the model and algorithm has been successfully used in Guangzhou railway bureau.

Performance Evaluation of Public Administrative Departments by Means of Benchmarking and DEA

WANG Qian, ZHOUZhuo-r

2004, 17(5): 585-589.

[Abstract](1234)

Abstract:
A new method was proposed by combining benchmarking and data envelop analysis (DEA) to evaluate the performances of public administrative departments. With the newmethod, the effective decision making units for a public administrative department are primarily determined with DEA. The primarily determined effective decision making units are combined with benchmarking units, and then the final effective decision making units are obtained by comparing more departments with DEA. The new method avoids subjective weighting in traditional evaluation of public administration units, and the disadvantage of relative effectiveness of DEA.

Synthetical Evaluation of Risks of Venture Business during Start-up Period

LIAOLi-kun, ZHANG Wei

2004, 17(5): 590-594.

[Abstract](1594)

Abstract:
To roundly understand and scientifically evaluate the risks of a venture business, a risk evaluation systemwith multiple evaluation indexes was set up based on deeply analyzing its characteristics and risk factors during a start-up period, and the fuzzy synthetical evaluation method was used to evaluate the risks of an enterprise. To overcome the shortcomings of the methods to subjectively or objectively determine the weight of indexes, the group eigenvalue method combining with the variation coefficient method was adopted. Finally, management risk was given to showthe application of the proposed method.

Optimization of Lot-Size for Dynamic Order with Quantity Discount

TIANJun-feng, YANGMei

2004, 17(5): 595-599.

[Abstract](1399)

Abstract:
To optimize the lot-sizes for dynamic orders with quantity discounts offered by vendors, a non- linear mixed integer programming model was proposed. In the model, multi-products, multi-discounts and capacities constraints of the purchaser and vendors were taken into account. The model was resolved with heuristic algorithm based on Tabu search technique. The computational results of an example validate the effectiveness of the model and the algorithm.

Vortex-Induced Dynamic Response of Tension Legs for Submerged Floating Tunnel under Current Effect

MAIJi-ting, LUO Zhong-xian, GUANBao-shu

2004, 17(5): 600-604.

[Abstract](1565)

Abstract:
Based on the equations of vortex-induced oscillation of tension legs, the effect of parametric excitation frequency on the first-order dynamic response of a tension legwas investigated by considering the non-linear fluid damping and parametric excitation. In the investigation, the Galerkin method and the numerical integration were used. Numerical relationships between displacement response, dynamic moment and dynamic shearing force and parametric excitation frequencywere obtained. The calculated result shows that the displacement response and the dynamic moment at mid-span both reach their maximums when the parametric excitation frequency is twice as large as the first-order natural frequency of a tension leg.

Numerical Simulation of Reinforcing Effect of Horizontal Drilling Jet-Grouting Pegs on Shenzhen Metro Tunnel between Stations

WUBo, GAO Bo, LUOJian-jun

2004, 17(5): 605-608.

[Abstract](1677)

Abstract:
To explain the reinforcing effect and mechanism of horizontal drilling jet-grouting pegs on surrounding rock around a tunnel, as an example, the reinforcing result of horizontal drilling jet-grouting pegs for Shenzhen metro tunnel between Dajuyuan and Kexueguan stations was analyzed with the 3D elastoplastic FEM (finite element method), and deformations induced by tunnel excavating with horizontal drilling jet-grouting pegs or not were compared. In the analysis, jet-grouting pegs were simplified as solid supportwith a thickness of 0.5 m. The numerical simulation result shows that the maximum excavation- induced settlements of vault and ground surface are reduced to 50% or so and plastic area in surrounding rock is reduced greatlywhen horizontal drilling jet-grouting pegs are used.

Patterns and Mechanisms of High Slope Deformation and Failure of Hydropower Station in Southwest China

SHI Yu-chuan, FENG Wen-kai, LIUHan-chao, ZHOUChun-hong, SHANZhi-gang

2004, 17(5): 609-613.

[Abstract](1293)

Abstract:
Based on systematic field survey and comprehensive analyses, the types, characteristics, influencing factors and mechanism of the deformation and failure of high slopes, located in two sides in the lock site region of a hydropowerstation in southwestChina, were investigated. The research shows thatthere are three types of the deformation and failure model of rock mass, they are sliding-fracturing, bending- fracturing, and flaking-peeling or fracturing-toppling. Therefore, there is a possibility for rock mass to generate the failures of falling and sliding, bending and toppling, and flaking and peeling or fracturing and toppling. Based on the above results, potential influence on the projectwas evaluated.

FEM-Based Parametrization Design of Cone Angle of Anchors

ZHAOTong, ZUO De-yuan, TONGJian-gang

2004, 17(5): 614-617.

[Abstract](1430)

Abstract:
A parametrical non-linear finite element model for mono-strand anchors was established. The influence of cone angle on the mechanical characteristics including deformation and stress distribution was analyzed, and a newwayto determine the optimum cone angle, parametrization design by usingFEM(finite elementmethod) was proposed. The research shows that when cone angle is less than some value, its increase will decrease the deformation of an anchor and raise its field load greatly, and for a mono-strand anchor, reasonable cone angle varies from 6.5°to 7.5°.

Random Optimal Vibration Control of Combined Isolation Structures

ZOULi-hua, ZHAO Ren-da

2004, 17(5): 618-622.

[Abstract](1432)

Abstract:
A new pattern of structures, combined isolation structure, was presented, and its vibration and control equations were derived. Based on the hypothesis that the input for earthquakes is white noise and using the principle of random optimal control, the features of vibration control and the effects of parameters such as isolation degree, damping and ground site on themwere analyzed. The research result indicates that the vibration control system of combined isolation structures can bring the seismic responses of structures under effective control by optimixing control parameters.

Experimental Research of Shear Lag Effect of Cable-Stayed Bridge withΠCross-Section Main Girder

WANZhen, LIQiao, MAOXue-ming

2004, 17(5): 623-627.

[Abstract](1607)

Abstract:
Amodel for a prestressed concrete cable-stayed bridge with aΠcross-section main girder was established in a scale of one to five for a concrete cable-stayed bridge over the Jinsha River in Yibin City. The shear lag effect of this bridgewas analyzed experimentally andwas comparedwith the one obtainedwith the finite element method. The comparison shows the reliability of the experiment. The experimental result shows that for a cable-stayed bridge with aΠcross-section main girder, shear lag factor is greater at the fixed ends of its cantilevers than at its mid-span.

Analysis of Loaded Frequency of Simply-Supported Beam Bridge

TANGHe-qiang, SHENRui-li

2004, 17(5): 628-832.

[Abstract](1582)

Abstract:
Based on the definition of natural bridge frequency, the analytical expression of loaded frequency of a simply-supported beam bridge under train loads was given through solving the vibration differential equation for bridges . The research shows that the loaded frequency of a bridge is in relation to the simplified models of a vehicle, the number of vehicles passing the bridge, train speed, bridge span and so on. The loaded frequency obtained is very proximate to the actual value when a vehicle is simplified as 4 or 2 wheelsets. It varies periodically when the total length of a train is larger than the bridge span. Furthermore, the loaded bridge frequency changeswithtime, relative tothe location ofvehicles on a bridge, and the faster a train moves, the more quickly the frequency varies.

Visual Load Paths in Elastic Structures Based on Equilibrium Principle

WANPeng, ZHENGKai-feng

2004, 17(5): 633-637.

[Abstract](1525)

Abstract:
Based on the definition of load paths, an approach was proposed to determine visual load paths in elastic structures. This approach is based on the equilibrium principle and combines the finite element method and the stream tracing algorithm. In order to get continuous visual load paths, commercial software Algor13 is used to get stress fields and finite element meshes and then import these information to another commercial software Techplot9.0 through data files. As an example, load paths in a cantilever beam were plotted by means of this approach and some differences between load paths and principal stress trajectories were compared.Finally, itwas pointed out that ideal load paths in an elastic structure are parallel andwell- distributed straight lines.

Bifurcations of 2-Multiple Semistable Limit Cycles of Airfoils with Cubic Nonlinearity

LIUFei, YANG Yi-ren

2004, 17(5): 638-641.

[Abstract](1344)

Abstract:
Bifurcations of 2-multiple semi-stable limit cycles, as well as supercritical and subcritical Hopf bifurcations of an airfoil flutter system with cubic nonlinearity in incompressible flows were studied. Air speed and the linear stiffness coefficient of pitching were taken to form a 2-dimensional parameter plane, and the analytic solutions of critical boundaries of Hopf bifurcations were obtained in the 2-dimensional parameter plane. As a result, the critical speed and linear stiffness for bifurcations of the 2-multiple semi- stable limit cycles were determined by means of harmonic balance method.

Bifurcation of Nonlinear Plate-Type Beams in Inviscid Fluids

LULi, YANG Yi-ren

2004, 17(5): 641-644.

[Abstract](1267)

Abstract:
The bifurcation of parallel plate-type beams with cubic nonlinear stiffness in inviscid fluids was investigated. A nonlinear model for a simply supported plate-type beam in inviscid fluids, a solid-liquid coupling system, was established on the basis of the hypothesis that all the plates have the same deflections at any instant. An algebraic criterion ofHopf bifurcationwas used to analyze the bifurcation ofthe structure. The result shows that there is no Hopf bifurcation for the structure in inviscid fluids. Finally, the static bifurcation of equilibrium positions and local stability of the structure were discussed. The numerical simulation shows that the coupling system is a nonlinear conservative system.

Synchronization between Two Chaotic Systems Using Lyapunov Method

TANXiao-hui, ZHANGJi-ye, YANG Yi-ren

2004, 17(5): 645-647.

[Abstract](1328)

Abstract:
Based on Lyapunov stabilitytheory and exemplified by aLorenz system, synchronization between two chaotic systems was achieved by directly transmitting the output of the drive system to the response systemto control the outputof the response system. Thismethod helps to avoid using complicated controllers and an easyway of synchronization between two chaotic systems.

Influence of Damping Characteristics of Dampers on Dynamic Performances of Four-Axle Power Cars

LUO Yun, CHENKang, JINDing-chang

2004, 17(5): 648-652.

[Abstract](1513)

Abstract:
A dynamic model for a four-axle power car was established and analyzed by using software SIMPACK and the nonlinear response analysis in time domain with random track irregularity to improve its dynamic performances. The numerical results show that the optimum damping coefficients of the secondary lateral damper, the primary and secondaryvertical dampers are often certain forthis kind of power cars, but the effect of the change in unloading force on the riding quality can be reduced by increasing unloading velocity to ensure its riding quality. For a four-axle power carwith secondary big deflection helical springs, it is important that the secondary lateral stop clearance matches well with the secondary lateral damping characteristic to obtain a good dynamic curve negotiating performance.

Establishment and Application of Prediction Model for Total Phosphorus in Channel Reservoirs

LIUYang, YANGLi-zhon

2004, 17(5): 653-655.

[Abstract](1424)

Abstract:
Based on the different hydrological and hydraulical conditions of channel reservoirs during a hydrological year, the prediction model of their total phosphorus (TP) was established by combining the 2- D straight river model and the polar coordinate diffusion model proposed by A B Karaueshev (A B Kapayщв). Then the TP prediction model was applied to simulating and forecasting the variation of TP concentration in narrow channel reservoirs with a water depth of less than 5 m. By taking the Sanjiang reservoir in Mianyang as an example, the relative error between predicted and observed results of its TP concentration is less than 10%.

SONG Wen； YILiang-zhong； MUXing-jun

SONG Wen, YILiang-zhong, MUXing-jun

2004, 17(5): 656-659.

[Abstract](987)

Abstract:
Based on the structural theory of Petri nets, the boundedness and liveness monotonicity of a subcategory of asymmetric choice (AC) nets, new extended strong asymmetric choice (NESAC) nets, were studied. The sufficient and necessary condition for determining structural liveness and boundedness of an NESAC net is that N is covered by a cluster of siphons and every nonemptyminimumsiphon must be a trap and satisfiest ∩H=t ∩H=1.NESAC net has the same characteristic of liveness monotonicity as ESAC net does.

Simulation Investigation of Channel Access Mechanism for 802.11e EDCF

ZHANGLian-bo, FANGXu-ming

2004, 17(5): 660-664.

[Abstract](991)

Abstract:
To deeply discuss the mathematics model for EDCF (enhanced distributed coordination function), the performances of two access mechanisms, 802.11 DCF (distributed coordination function) and 802.11e EDCF, were analyzed and compared through simulation. The differences between 802.11 DCF and 802.11e EDCFwere discussed, and the principle of EDCF providing the quality of service (QoS) was investigated. Simulation result shows that EDCF can provide the guarantee of QoS forwireless LANs (local area network), but systemperformance deteriorates if there are many data streamswith high priority contend for a channel.

A Family of Irregular LDPC Codes Based on Optical Orthogonal Codes

WENHong, HUFei, JINFan, DUANHui-yong

2004, 17(5): 665-669.

[Abstract](1081)

Abstract:
Based on the optical orthogonal codes, a method for constructing irregular LDPC (low-density parity check) codes was presented. These LDPC codes are quasi-cyclic codes and can be encoded with low complexity with a linear relationship to code length. These codes haveTanner graphs free of 4-cycles. They performwellwith the sum-product iterative decoding. Comparedwith random codes and regularLDPC codes based on the optical orthogonal codes with similar parameters, they have decoding performance gains respectively about 0.3 and 0.15 dB when bit error rate is 10-5.

Automatic Abstraction Algorithm for Timed Automata

ZHIXiao-li, TONG Wei-qin, RONGLu

2004, 17(5): 670-674.

[Abstract](1264)

Abstract:
To deal with the state-explosion problem in the specification analysis and verification of timed automata, an automatic abstraction algorithm for timed automata was proposed. By directly abstracting the specifications of a timed automaton instead of its semantic model ofmuch bigger size, this algorithmyields a polynomial time complexity no more than the four power of its topology complexity. An experiment shows that the algorithm can be applied to any timed automaton satisfying a property called linear resetting and used to simplify system specifications to improve their understandability.

New Comprehensive Learning Method Based on LM-QuasiNewton Algorithm Applying for Feed-Forward Neural Network

XUJin

2004, 17(5): 675-679.

[Abstract](1454)

Abstract:
To solve the large residual problems which may occur during weight training, QuasiNewton algorithm was combined with Levernberg-Marquardt algorithm to form a new comprehensive learning algorithm, named LM-QuasiNewton, for feed-forward neural networks. Simulation shows that LM- QuasiNewton algorithm can effectively solve large residual problems with better convergence and stability compared with otherweight learning algorithms.

Asynchronous Parallel Iterative Algorithm Solving Equation of Higher Degree

YANG Ben-li, LIAn-zhi, ZENGXian-wen, HANWei-hua

2004, 17(5): 679-683.

[Abstract](1299)

Abstract:
By using the method for positive term resolution of equations of higher degree, all non-zero real roots of a real coefficient equation of higher degreewere obtained by determiningthe abscissas of intersection points of two monotonically increasing concave functions in the first quadrant of a planar rectangular coordinate system. An asynchronous parallel iterative algorithm based on the shared memory MIMD (multiple instruction and multiple data streams) parallel computation model was put forward. This iterative algorithm has the characteristic of global convergence and can be used to determine all the real roots of any real coefficient equation of higher degree. In addition, the complexity of the algorithmwas discussed.

Properties and Computation of Linear System Transfer MatrixH∞-norm

XUYue-liang, LI Zhi

2004, 17(5): 684-687.

[Abstract](1477)

Abstract:
An equivalence theorem was derived to determine if the transfer matrixH∞-norm between two linear systems [A,B,C,D] and [A+BR-1DTC,BR-1/2,(I+DR-1DT)1/2C,0] is smaller than 1. This theoremwas then used to derive another theorem about the stability of system [A,B,C,D]. This equivalence theorem simplifies the proof of the theorem about the stability of system [A,B,C,D] and determination ofH∞-norm. An algorithm for computing theH∞-norm of the transfer matrix of system[A, B,C,D] was presented.

Interval Estimation of Chance-Constrained Programming Based on Genetic Algorithm

CHENGao-bo, LIUHai-yan

2004, 17(5): 687-690.

[Abstract](1817)

Abstract:
The interval of the optimal value of the non-linear objective function of chance-constrained programmingwas estimated from the viewpoint of mathematical statistics, and the ways of improving the precision of the estimated interval were discussed. The genetic algorithm and the spline regression were adopted to obtain the Lipschitz constant of the fitted optimum objective function. A calculational example shows the validity of the interval estimation for the optimal value of a non-linear objective function.

Semantics of Lattice-Valued Tense Propositional Logic System

LI Wen-jiang, XUYang

2004, 17(5): 691-695.

[Abstract](1261)

Abstract:
Tense operatorsE(ever)andF(will)as well as their dual operatorsH(ever always be) andG (will always be) were introduced into lattice-valued propositional logic system LP(X), forming a lattice- valued tense propositional logic system LTP(X).It takes time axis as language circumstances. The properties of LTP(X)related to time and superposition of tense words were discussed. It was proved that semantic modus ponus (MP) and hypothetical syllogism(HS) rules hold in this logic system.

Comparison of Dissolving Processes for Preparation of Nano-CeO2Powers

ZHANGXi-yan, GAOJie-ming, ZHOUShi-jie, ZHANGJin

2004, 17(5): 696-698.

[Abstract](1081)

Abstract:
The dissolving processes with HCl and HNO3were compared experimentally. The proportion of dissolved CeO2with HCl was less than 25% accompanied with large consumption in raw materials and serious pollution by Cl2, unsuitable for production. On the other hand, the proportion of CeO2dissolved withHNO3was over 90% in the following reaction conditions: temperature 35 to 50℃, concentration of HNO3: 7.0 to 8.5 mol/L,VH2O2:VHNO3=0.125, dissolving time: 60 min. The resolved nano-powders were homogeneous and theirgrain sizeswere in the range of 30 to 40 nm. HNO3dissolving process saves raw materials and there is no pollution, showing industrialized production.

Research of Reflection Characteristics of Magnetic Boundary

GUANZheng-tao, ZHANG Shuang-wen, LIUYun-lin, YANG Ru-gui

2004, 17(5): 699-702.

[Abstract](1190)

Abstract:
To provide a basis for the deep research of the transmission characteristic of magnetic radome, the reflection characteristics of an air-magnetic boundaryon a planarwavewith different incident angleswere investigated. By changing the parameters of magnetic media, the relationships between the reflection coefficient of power as well as the phase of reflection coefficient at the boundary and the incident angle was analyzed, and the effects of the magnetic property of media on the reflection coefficient and the phase were discussed. The research resultshows that a planarwave goingthrough amagnetic boundary does not produce the depolarization effect when the relative complex permittivity of the magnetic medium equals its relative complex permeability.

Search Advance Search

Journal Search

Search Type

NewsMore

instruction for authors

Read this journal

LinksMore

2004 Vol. 17, No. 5