西南交通大学学报

Simple Method of Determining the Initial Value in Secondary Exponential Smoothing Method

WANG Ci-guang

2004, 17(3): 269-271.

[Abstract](1417)

Abstract:
When the secondary exponential smoothing is applied to a trend prediction model, its initial value has to be determined,which makes the process complex and time consuming. Through derivations, it was proved a simple method to utilize the initial index value as the smoothing one. The feasibility and applicable conditions of the method were presented.

Chance-Constrained Shadow Prices

PENG Yu, JIA Zhi-yong

2004, 17(3): 272-276.

[Abstract](1418)

Abstract:
To promote researches on stochastic programming (SP), some characteristics of SP and its dual programming were obtained via defining the dual programming of reliant programming for chance-constrained programming. From the optimal solution to the dual programming, the concept of chance-constrained shadow prices was proposed. It reflects the influences of constrains to the optimal values of goal functions. An example was presented to show the application of chance- constrained shadow prices in economic decision making.

Multidimensional Grey Evaluation Model for Suppliers Selection and Evaluation

SUN Chao-yuan, PENG Qi-yuan

2004, 17(3): 277-280.

[Abstract](1305)

Abstract:
Multidimensional grey evaluation method was used to select suppliers from four indexes such as price, quality, delivery period and service level. This method can avoid the evaluation error due to subjective factors in a certain degree and obtain full information about price, quality, delivery period and service level. An example was presented to prove the feasiblity of the proposed method.

Modified Clark-Wright Algorithm for Vehicle Scheduling Problem with Fuzzy Cost Coefficients

ZHANG Jian-yong, GUO Yao-huang, LIJun

2004, 17(3): 281-285.

[Abstract](1402)

Abstract:
The traditional deterministic vehicle scheduling problem (VSP) was expanded into a VSP with fuzzy features. The cost coefficients of the VSP with the features were treated as fuzzy numbers. After a simple description of the VSP with fuzzy cost coefficients, a mathematical model for the fuzzy scheduling of vehicles was set up. And a modified Clark-Wright algorithm for this kind of vehicle scheduling problem was proposed based on the effective combination of the classical Clark- Wright algorithm and the Gaufmann-Gupta method for ordering of fuzzy numbers. Finally, an example for this problem was given.

Action of International Direct Investment on Invested Country Economics

HUANG Jian-hong, PU Yun

2004, 17(3): 285-290.

[Abstract](1449)

Abstract:
In order to better attract and utilize foreign capital, non-linear programming was used to investigate the action mechanism of international direct investment and its properties from the viewpoint of labor and specialization divisions. It was pointed out that international direct investment is not always favorable for the economic development of an invested country, and improving the transaction efficiency related to institution plays an important role in attracting foreign capital. Moreover, to an underdeveloped region, to invest foreign capital in the infrastructure that can improve transaction efficiency, instead of other industries, is more helpful for its economic development.

Solution of Complete Style of Interval Linear Programming

LI Wen-hua, GUO Jun-peng

2004, 17(3): 291-295.

[Abstract](1231)

Abstract:
Based on the current related researches, a complete style of interval linear programming was defined. All the coefficients of its objective function and constraints can be interval numbers, and its constraint conditions may be equations or inequations. An order relation between interval numbers reflecting the satisfactory level of decision makers was proposed, and based on this, inequality constraints of interval linear programming were transformed into constraints with exact coefficients. The equality constraints of interval linear programming were discussed and converted into inequality constraints with exact coefficients. In addition, its objective function concluding interval numbers was analyzed and changed into an exact objective function. As a result, an interval linear programming problem was transformed into an exact linear programming problem and solved. Finally, an example was given.

Gradually-Vanishing Asset and Its Pricing Model

ZHANG Xiu-min, HU Yang, RANHui-juan

2004, 17(3): 296-300.

[Abstract](1074)

Abstract:
In view of the characteristic of some assets different from perished assets,i.e., their values vanish gradually, conception of gradually-vanishing asset (GVA) was put forward. From the relationship between the number of products and the need of consumers, they were divided into two classifications. For one, the need of consumers can be met in most cases, and for the other, supply and demand do not match, with a seasonal or periodical change. A mathematical expression was proposed for the first kind of GVA, and a new pricing method—gradually-vanishing pricing model was brought forward. The result based on the comparison with other pricing methods shows that this method can decrease GVA value elapsing, raise enterprise revenue, increase consumer remnant, and further improve social welfare.

Nonlinear Programming Model and Algorithm of Two-Layer Distribution Network Based on Integrated Transit-Inventory Concept

ZHANG Jin, MA Xiao-lai, DU Wen

2004, 17(3): 301-305.

[Abstract](1432)

Abstract:
Based on analyzing the existing models of transit and inventory and the characters of a modern logistics distribution network, a nonlinear two-layer distribution network model was put forward from an integrated transit-inventory concept, and a non-linear programming model, with the restriction of capacity and requirement and the objective function of transport cost and distance, was proposed. Furthermore, its optimum solution was proved, and the calculational process for the algorithm was given. Finally, the algorithm was verified by using an actual example. The result shows that the research is useful to establishing the model of multi-phase, multi-line and multi- target logistics distribution.

Hardware Design of Fast Module Exponential Algorithm

FAN Ming-yu, WANG Jian-hua, WANG Guang-wei

2004, 17(3): 306-310.

[Abstract](2059)

Abstract:
Fast module exponential algorithm, fast module multiplication algorithm and fast module addition algorithm were integrated to form a hybrid fast RSA algorithm for large number operations. A hardware scheme for implementation of RSA key was proposed using the proposed algorithm. In the algorithm,m-ary method is used in the first stage to reduce multiplication times, then modified Montgomery method is used to decrease addition times, and finally fast adders and multiplication unit are made to work synchronously by adjusting their structures. A processing speed of 390 kbit/ s at 100 MHz clock was achieved for a 1 024 bit key operand.

Resonant Frequency of Cavity-Backed Rectangular Patch Antenna with Metal Column

LIU Yun-lin

2004, 17(3): 311-313.

[Abstract](1299)

Abstract:
A finite element-boundary integral formulation was introduced to investigate the resonant frequency of a cavity-backed rectangular patch antenna with a metal column. The effects of the size of the metal column and the position of feeding points on the resonant frequency were discussed in detail. The research results show that the resonant frequency of the cavity-backed rectangular patch antenna rises with the size increasing of the metal column, and its inner field is different from that of a common patch antenna.

Improved Beamforming Algorithm Based on Subspace Tracking

WANG Jian-ying, DENG Ya-bo, YIN Zhong-ke, CHEN Lei

2004, 17(3): 314-317.

[Abstract](1364)

Abstract:
Based on the ESPRIT (estimating signal parameter via rotational invariance techniques) method, an improved algorithm for beamforming was proposed by utilizing the subspace tracking algorithm in perturbation analyses to obtain ideal beamforming under the condition that the direction of expected signals is unknown or changed. In this algorithm, real-time tracking for the direction vector of the expected signals is realized due to the application of the subspace tracking algorithm, as a result, the pointing error of time-varying signals can be controlled. Numerical simulations show that beamforming obtained using the improved algorithm approaches to the beamforming under the ideal condition that the direction of the expected signals is known.

Application of Multiwavelet Multiresolution Analysis in Signal Processing of Power Systems

LIUZhi-gang, ZENG Yi-da, QIAN Qing-quan

2004, 17(3): 318-323.

[Abstract](1445)

Abstract:
Multiresolution analysis method of multiwavelet was briefly introduced. Its application in fault monitoring, fault data compression and denoising was studied, and its performances in fault diagnosis and monitoring and fault data processing were compared with traditional wavelet. Simulation results show that the performances of multiwavelet is better in general than those of traditional wavelet if a proper preprocessing method is adopted in multiwavelet.

Optimal Design of Wavelength-Convertible Optical Switch for Optical Networks

ZHANG Lan

2004, 17(3): 324-328.

[Abstract](1331)

Abstract:
In order to resolve the contradiction between the high cost of wavelength converters and their critical roles in optical networks, a new architecture of wavelength-convertible switches was put forward, and its optimal routing and wavelength assignment algorithm was proposed. Simulations were done on the NSF network to justify the advantages of the proposed architecture. Moreover, the optimal number of wavelength converters and their correct arrangement at switching nodes were obtained through the simulations. The proposed switch architecture along with the simulation tool can be effectively applied to other long-haul optical networks.

Performance Comparison of Local Zero-Order Predictive Methods for Chaotic Time Series

LIHeng-chao, ZHANG Jia-shu

2004, 17(3): 328-331.

[Abstract](1716)

Abstract:
Performances of local zero-order prediction methods for chaotic time series were compared in aspects of prediction accuracy, anti-noise and the ability of multi-step prediction through computer simulation. Simulation results show that distance-weighted predictive method is the best when there are no noises or only weak noises; exponential weighted predictive method is better than the others when there are large noises; exponential weighted predictive method and averaging method are basically the same in respect to multi-step prediction, but distance-weighted predictive method is the best in short-term prediction; multi-step prediction errors of standard discrete chaotic time series for the three methods do not increase with prediction step after they arrive at certain values, but for the chaotic time series sampled by continuous system, its multi-step prediction errors have periodicity.

Fuzzy Logic-Based Direct Torque Control of Permanent Magnetic Synchronous Motor

WANG Jun, PENG Hong

2004, 17(3): 332-336.

[Abstract](1369)

Abstract:
In order to improve the system stability and raise the response speed, a new control scheme, fuzzy logic-based direct torque control, for a permanent magnet synchronous motor (PMSM) was put forward. In this scheme, zero voltage space vectors are introduced to simplify fuzzy control rulers. Simulation results show that compared with the common direct torque control, the flux locus of the fuzzy direct torque control is more approximatively a circle, and the rotational speed can reach up to a stable value quickly. Subjected to an external disturbance, the torque for the proposed scheme can change smoothly and reach to a new stable value. Furthermore, the high frequency components of the stable current decrease greatly.

Information Fusion and Cluster Control System for Natural Gas Flow

YU Li-jian, YANG Hua

2004, 17(3): 337-340.

[Abstract](1615)

Abstract:
According to the mathematical model of natural gas flow, a flow control system was proposed based on information fusion and cluster analysis. The system consists of an ART-2 network and a BP network. Pressure differences, pressures and temperatures obtained by sensors are fused and the output spaces are controled according to cluster control strategy. This system has been used in a romote monitering and control system. The operation of the system shows that the pressure fluctuation is controled within 10% when the range of the natural gas flow is between 5% and 95 % of the maximum flow.

Scheduling of Reconfigurable Manufacturing System Based on Timed Reconfigurable Petri Net

CAI Zong-yan, WANG Ning-sheng, REN Shou-gang, LI Ya-bai

2004, 17(3): 341-344.

[Abstract](1858)

Abstract:
The proposed scheduling algorithm for reconfigurable manufacturing systems is based on timed reconfigurable Petri net. It generates and searches a partial reachable graph to find an optimal or near optimal feasible schedule according to the firing sequence of transitions of the timed reconfigurable Petri net model for a reasonable computational cost. A scheduling example was presented to demonstrate the correctness and feasibility of the algorithm.

Defect Detecting of Train Wheelset Tread Surface with Laser Displacement Sensor

AO Yin-hui, XUXiao-dong, WUNai-you

2004, 17(3): 345-348.

[Abstract](1487)

Abstract:
A novel method for automatically detecting the scrape and exfoliation of train wheelset tread surfaces was proposed. In this method, data on train wheelset tread surfaces are gained by using laser displacement sensors, and analyzed through deviation rectifying, interpolation and multistage median filtering. The digital matrix for the tread surfaces is digitized so as to gain their status. A universal software for the measuring system was developed with VC++ and Matlab programming languages. This software can automatically show 2-and 3-D views of train wheelset tread surfaces, and print the image of scrape and exfoliation areas.

Aposteriori Detection of Model Error in Surveying Adjustment

FAN Dong-ming

2004, 17(3): 349-352.

[Abstract](1262)

Abstract:
A parameter elimination method to convert a parameter adjustment model to a condition adjustment model was brought forward. Using this method, a parameter adjustment model can be changed into a condition adjustment model through eliminating all the parameters in the parameter adjustment model. As a result, condition equations can be automatically established with computers. Based on the above research, the approach of outlier detecting and locating with the closing errors of condition equations was given. In addition, the feasibility of this method was proved by an example. The result shows that this method can test all the checkout conditions including side and coordinate conditions, raise the capability of outlier detecting and locating in an adjustment system, and make the searching of closed loops in a surveying network realized satisfactorily with computers.

Spatial Analysis of Traffic Geographical Information

XUJing-hua

2004, 17(3): 353-357.

[Abstract](1433)

Abstract:
Based on the methods and techniques of geographical information system (GIS) and traffic geographical information system (GIS-T), the main spatial (configuration) characteristics and geographical background information of a traffic network were analyzed from itself and its background factors and by combining an experiment on the related data for Chengdu City using MapInfo, and the method and technique of the analyses were discussed. The research and experiment show that the method of spatial analyses for traffic geographical information based on GIS has more advantages over the traditional analysis methods in time, effect, accuracy, and integration and so on.

Urban Spatial Visualization Based on Principle of Sub-water Sight

FENG De-jun, LI Yong-shu

2004, 17(3): 358-361.

[Abstract](1348)

Abstract:
Based on the principle of sub-water sight, the 3D model of urban overground and underground objects was set up in order to solve the problem of visual confusion resulted from modeling directly from their 3D coordinates. According to the principle, when the vision point is located on the ground, a proper color can be given to the earth s surface by using the color blend function of OpenGL, and underground objects are made to have the color ingredient of the earth s surface to distinguish overground and underground objects from their colors. On the condition of great underground depth and more underground objects, information on their depth can be shown by the shading degree of their colors by means of the fog operation function of OpenGL. At last, the feasibility of this method was proved using data on an urban district.

Experiments of Support Structure of Twin-Bore Highway Tunnels

LIN Gang, HE Chuan

2004, 17(3): 362-365.

[Abstract](1677)

Abstract:
Similar simulation tests were made to determine the reasonable structure and principle of supporting for twin-bore highway tunnels with four driveways excavated in classⅡsurrounding rock, and in-situ testing was carried out to prove and complement the experimental result in laboratory. The research results show that because stresses in the mid-partition of twin-bore highway tunnels become very complexly in the construction process, main attention shall be paid to preventing the mid-partition from turning over during the construction. In addition, rock bolts and shotcrete, instead of steel supports, should be taken as the main preliminary supporting of twin-bore highway tunnels with four driveways excavated in classⅡsurrounding rock. Finally, the principle of supporting was put forward for twin-bore highway tunnels with four driveways excavated in classes ⅠandⅡsurrounding rocks.

Mechanism of Sandpile Avalanche in Critical State

LI Shi-xiong, YAO Ling-kan, JANG Liang-wei

2004, 17(3): 366-370.

[Abstract](1288)

Abstract:
The dynamical characters of granular mixtures were investigated based on the concept of self-organized criticality (SOC). Through sandpile experiments with one-sided slope, it was discovered that a sandpile consisting of non-uniform sand with a sorting coefficientφ=2.85 exhibits SOC, while the large-scale avalanches of a sandpile composed of uniform sand with a sorting coefficientφ≈1 have a quasi-periodicity. It was believed that sandpiles have representative binary composition,i.e.a granular layer and a friction-resistance layer. The result indicates that in the critical state large-scale avalanches happen mainly in the granular layer for a non-uniform sandpile, whereas in both the granular and friction-resistance layers for a uniform sandpile. This shows that the dynamic characters of granular mixtures in the critical state are controlled by the non-uniform degree, and large-scale avalanches have two possible forms for sandpiles with different non-uniform degrees,i.e.,avalanche in the granular layer and avalanche in the granular and friction-resistance layers.

Investigation on Local Stresses in Anchorage Zone of Continuous Rigid Frame Bridge

XU Wei-guo, HE Guang-han

2004, 17(3): 371-374.

[Abstract](1555)

Abstract:
To arrange the prestressed tendons and anchor devices on the web of box girder conveniently, a theoretical analysis, numerical calculation and model test on local stresses in concrete under anchor devices were performed separately using a typical segmental model, a subrange model and a similar model for the scheme design of Hutiaomen major bridge. The results show that higher concrete stresses occur only in a very small zone under anchor devices, and large prestressing force designed in light of the real bridge does not cause extremely high principal stress and concrete failure. In addition, the measured values of local stresses in concrete in an anchorage zone are comparatively close to their calculated values.

Nonlinear Time Domain Buffeting Response Analysis for Long-Span Railway Cable-Stayed Bridge

LI Yong-le, LIAO Hai-li, QIANG Shi-zhong

2004, 17(3): 375-380.

[Abstract](1756)

Abstract:
Based on the time domain wind loads obtained by wind field simulation, the nonlinear behaviors of bridge buffeting were investigated by the buffeting analysis method for long-span bridges. In the investigation, a long-span railway cable-stayed bridge was regarded as an example, and both structural geometric nonlinearity and aerodynamic nonlinear factors were taken into account simultaneously. The geometric nonlinear factors include cable sag, beam-column effect and large structure displacement, and the aerodynamic nonlinear factors are wind incidence angle effect and self-excited force. The nonlinear kinetic equation of a bridge structure was solved with the double iteration method. The comparison between linear and nonlinear analysis results shows that the nonlinear factors can result in a great buffeting response of long-span railway cable-stayed bridges.

Experimental Research on Separation of Concrete-Filled Steel Tube and Effect of Pouring Pulp

YE Yue-zhong, WEN Zhi-hong, PAN Shao-wei

2004, 17(3): 381-385.

[Abstract](1382)

Abstract:
The mechanical properties of concrete-filled steel tubes (CFST), including elastic modulus and load-bearing capacity etc, and the effect of pouring pulp were investigated experimentally under the condition of separation of concrete and steel tube. The result indicates that with the increase of separation width, the elastic modulus and load-bearing capacity of a CFST decrease greatly, and its longitudinal and lateral deformations increase sharply. The reason for this is that when the separation of concrete and steel tube happens, they can not bear loads together. After pulp is poured for a separated CFST, it can renew its original mechanical properties basically.

Graph-Based Algorithm for Mining Frequent Closed Itemsets

LI Li, ZHAI Dong-hai, JIN Fan

2004, 17(3): 385-389.

[Abstract](1368)

Abstract:
To raise the efficiency of data mining, a graph-based algorithm for mining frequent closed itemsets, called GFCG (graph-based frequent closed itemset generation) was proposed. In this algorithm, the bit vector technique is used to construct a directed graph to represent the frequent relationship between items, and frequent closed itemsets are generated recursively from this graph. As a result, the GFCG scans the database for only two times, and does not generate candidate sets. Furthermore, the concept of an expanded frequent itemset is introduced to greatly decrease the searching range for adjusting whether a frequent itemset is closed or not. In addition, this algorithm was tested by using one actual and two synthetical databases. The testing result shows that compared with the A-close and CLOSET algorithms, the proposed algorithm has good speed and scale-up properties.

Quantum Evolution Algorithm Based on Descending Search

MA Shu-xia

2004, 17(3): 390-393.

[Abstract](1747)

Abstract:
To raise global search capacity and convergent speed, a new evolution algorithm, based- descending search quantum evolution algorithm, was put forward on the basis of the quantum evolution algorithm (QEA) and the hybrid genetic algorithm. In the proposed algorithm, the descending search theory of optimization principles is applied, so the iterative effect, only relying on quantum gate, of QEA is improved to speed up the convergent speed, and the possibility of individual retrogression in the evolution process is reduced. The simulation result of a typical function shows that this algorithm has a good convergence performance and global search capacity.

Remarks on Gronwall Inequality

LIANG Shao-jun

2004, 17(3): 394-396.

[Abstract](933)

Abstract:
The classical Gronwall inequality and its proof by using general methods were introduced. The estimate of linear control was extended to that of nonlinear control by the comparison theorem. The application of Gronwall inequality to the study of the characteristics of solutions of differential equations was also discussed.

Spectrum of Polynomial Conjugate OperatorT

CUIHong-wei, HAN Liu-bing

2004, 17(3): 397-399.

[Abstract](1204)

Abstract:
The concept of a special normal operator, self-conjugate operator, in Hilbert space was extended to a polynomial conjugate operator. Based on the properties and spectrum feature of a normal operator in Hilbert space, the spectrum of the extended operator—polynomial conjugate operatorTwas investigated using the analogy method and the Banach algebra technique. The result shows that the spectrum of the polynomial conjugate operator is a finite eigenvalue spectrum. The conclusion was proved by using an example

Extension of Lin-Bairstow Method

FENG Tian-xiang

2004, 17(3): 400-402.

[Abstract](1048)

Abstract:
Lin-Bairstow method was extended to derive the factors of polynomials with more than 3rd-degree. A numerical method for solution ofkth-degree (k≥3) factors of real coefficient polynomials was proposed. An approach similar to Newtonian iteration method is used in the proposed method, and the error is controlled by the error from the process of finding the roots with Newtonian method

Matrix Solution of Integral Indeterminate Equations of First Degree and Its Program Design

YAN Lin

2004, 17(3): 403-407.

[Abstract](1401)

Abstract:
By using the Euclidean algorithm and invertible linear transformation over an integral ring, the solution of integral indeterminate equations of the first degree was investigated in theory, and its matrix solution based on the elementary matrix transformation was proposed. In addition, the program for the matrix solution was given by using MATLAB mathematical software.

Blow-up of Solution of Nonlinear Evolution Equation

YANG Zhi-feng, HAN Liu-bing

2004, 17(3): 408-410.

[Abstract](1227)

Abstract:
Based on the energy conversation law for a nonlinear evolution equation, blow-up of the initial value of its solution was investigated by using the improved convexity analysis method and the Sobolev s embedding theorem. A sufficient condition for blow-up of the initial solution of this equation was obtained, that is, the initial energyE(0)has a certain upper bound. And this upper bound is only related to the Sobolev s embedding constant in the discussed space.

Study on Karyotypes of SixRanunculusSpecies in Sichuan

LUO Hong-xia

2004, 17(3): 411-414.

[Abstract](1076)

Abstract:
The somatic chromosome number and detailed chromosome morphology of six Ranunculusspecies in Sichuan were studied. The results show that their chromosome numbers are 2n=16 forR.tanguticus, 2n=32 forR.nephelogenes, 2n=16 forR.ficariifolius, 2n=32 for R.sceleratus, 2n=14 forR.japonicus, and 2n=16 forR.silerifolius. From these chromosomal data, the genetical relationships of four sections inRanunculuswere discussed. The conclusion is that sect.Ranunculus among the four sections shows an advanced level in evolution.

Synthesis of Bis-azo Chromogenic Acyclic Crown Ethers

TONG Zhi-ping, LIUJing, GONG Zhong-lun

2004, 17(3): 415-418.

[Abstract](1014)

Abstract:
Two new types of bis-azo chromogenic acyclic crown ethers with brigded chains of azachains and terminals of phenolic hydroxyl and chromogens of 4-R benzenediazo groups 5 and 6 have been conveniently synthesized with good yields via coupling-bridging reactions from salicyladehyde, 4-nitro (-chloro,-methoxy) benzeneamine, ethyldiamine and diethylenetriamine, respectively.Their structures of them have been confirmed by MS,IR measurement and elemental analyses.

Search Advance Search

Journal Search

Search Type

NewsMore

instruction for authors

Read this journal

LinksMore

2004 Vol. 17, No. 3