2010 Vol. 36, No. 4

Display Method:
Cognition behavior model for air combat based on reinforcement learning
Ma Yaofei, Gong Guanghong, Peng Xiaoyuan
2010, 36(4): 379-383.
Abstract:
A cognition model was proposed to support tactical decisions for simulated fighters to fight with each other in a virtual combat, and reinforcement learning (RL) technology was used to acquire knowledge. The combat situation was described by multi-attributes, which resulted in a high dimensional problem space in which the fighters learned to find action policies. The traditional approach that partitioned the problem space would impose demand on huge computation and storage resource. An approximation network is constructed based on Gaussian radial basis function to approximate the state value, which greatly reduced the resource demand and learning cycle time, and produced reasonable maneuver strategy. The model was verified by a one-to-one air combat simulation, and the produced trajectories are similar with those that human pilots flied in real combat.
Deformation behavior of sands under cyclic non-proportional loadings
Tong Zhaoxia, Zhang Jianmin, Yu Yilin, Zhang Ga
2010, 36(4): 384-387.
Abstract:
Drained tests for non-proportional loadings involving with cyclic variations of the deviatoric stress and principal stress axes were conducted on sands with a hollow cylinder apparatus at Hong Kong University of Science and Technology. The test condition was controlled to be that principal stress axes rotated cyclically from 0° to 180° to the vertical axis accompanied by the variation of the deviatoric stress simultaneously. The developments of three normal strain components and the volumetric strain with the number of cycles and the relationship between the shear stress and the shear strain components were analyzed in detail. With the increase of the cycle number, the contractive volumetric strain of sands induced is increasing, yet its increase rate is lowered. And it is found that the magnitude variation of the deviatoric stress plays a significant role in the deformation behavior of sands under cyclic non-proportional loadings.
Numerical simulation of two-dimensional fluidic throat skewing vector nozzle
Wang Fei, Eriqitai Li, Jiajun Wang
2010, 36(4): 388-390.
Abstract:
The flow fields of the two-dimensional fluidic throat skewing vector nozzle were simulated numerically. The effects of the single side injection at the throat and the assisted injection at the divergent section on the flow and the performance of the nozzle were studied. These simulations show that the single side injection at the throat can generate an asymmetric flow, and result in thrust vector, but the vector angle is smaller; the assisted injection at the divergent section can improve the nozzle thrust performance remarkably; only with the larger injection flux, there will be a typical "throat skewing" phenomenon; however, the highest efficiency region of thrust vector control is not occurred after the "throat skewing" phenomenon, but in the process of the forward movement of the bow shock position and the expansion of the subsonic region upstream of the divergent section injection inlet.
Model-based optimization and calibration of piston aero-engine
Zhao Lifeng, Li Yunqing, Cheng Chuansong, Wang Haiying
2010, 36(4): 391-394.
Abstract:
Multiobjective optimization model of aero-engine torque and fuel economy was established in Boost and Matlab environment. The progress of the calibration was elaborated, including building Boost model, design of experiment(DOE)tests, mathematical statistical modeling ,the response surface of variables was explored to the relationship between the objectives and variables, the statistical modeling can control errors. The model-based calibration(MBC) toolbox was applied to optimize the model, the method of normal boundary intersection is economical and applicable to deal with multiobjective optimization of calibration. The map of ignition timing and the amount of injection fuel in each cycle were given, the relationship between the target variables and independent variables was given with response surface map, which are useful to improve the performance of aero-engine.
Large mass flowrate gas-gas injector optimization design with response surface methodology
Du Zhenggang, Li Mao, Jin Ping, Cai Guobiao
2010, 36(4): 395-398.
Abstract:
In order to investigate the effect of design parameters of dual shear coaxial injector that utilizes gaseous oxygen and hydrogen as propellants on the combustion location, the numerical simulation efforts were performed to simulate the flow field of combustor with orthogonal array applied to guide the injector design. The results display that dual shear coaxial injector made propellants combusted efficiently at the condition of large mass flowrate which was close to eight times greater than that of space shuttle main engine (SSME) single element injector. The range analysis and variance analysis indicate that the velocity ratio of hydrogen to oxygen has significant effect on the combustion location, while the ratio of center hydrogen mass fraction and oxygen wall thickness have not influenced the combustion location significantly. The response surface methodology based on orthogonal-polynomial is employed to get the optimized parameters of the dual shear coaxial injector.
Singularity free universal variables method to solve Lambert transfer
Peng Kun, Xu Shijie
2010, 36(4): 399-402.
Abstract:
The conventional universal variables method has singularity in solving the velocities at transfer time of Lambert transfer orbit when angle between two position vectors is 180°, and may lose some important universal variables by using Newton method. The singularity free formulas for solving velocities were derived by using formula transformation to expurgate singular terms. And the range of universal variables was divided into several sections according to the loop number of transfer orbit, then artificial immune algorithm(AIA) was used to search universal variables solutions in each section, finally all Lambert transfer orbit were derived and the optimal one can be chosen. Simulation results show that, this modified universal variables method can availably avoid the singularity of conventional universal variables method, derive all Lambert transfer orbit and get the transfer orbit whose fuel penalty is optimal.
Pilot performance evaluation model based on BP neural networks
Liu Zhongqi, Yuan Xiugan, Fan Yubo
2010, 36(4): 403-406.
Abstract:
In order to evaluate pilot performance objectively, back propagation(BP) neural network model of 6-14-3 form in topology with eye movement data was established. Data source of BP neural networks that came from former experiment and random interpolation was divided into training set and test set and normalized. Based on neural networks toolbox in Matlab, hidden layer nodes of BP networks were determined with empirical formula and experimental comparision; BP algorithms in the toolbox were optimized; The training set data and test data were input into model for training and simulation; Pilot performance of the three skill levels was predicated and evaluated. The research shows that pilot performance can be accurately evaluated by setting up BP neural networks model with eye movement data and the evaluation method can provide a reference for flight training.
Class of multiple-revolution impulsive rendezvous with priority of minimum fuel
ZHANG Feng-Yu, YANG Meng, WANG Song-Yan, WANG Zi-Cai
2010, 36(4): 407-410.
Abstract:
For two-impulsive rendezvous using multiple-revolution Lambert solutions, the relationship between characteristic velocity (Δv) and semi-major axis of transfer trajectory was considered. It was proposed that the optimal solution actually is the less fuel trajectory among 2N+1 trajectories satisfying time constraint, but not the minimum fuel trajectory(MFT). As the first impulse of two-impulsive rendezvous was dissembled into two impulses with the same direction, a chaser could consume the redundant transfer time by coasting N revolutions on a specified orbit, and rendezvous a target on MFT in the rest transfer time. It was proved that Δv of this rendezvous coincide with that of minimum fuel transfer in geometry. The existence of solutions was given. Some simulations show that this rendezvous can save fuel and the existence of solutions is more loosely restrictive on the length of transfer time than two-impulsive rendezvous using multiple-revolution Lambert solutions.
Structural dynamic response characteristics of supercavitating underwater vehicles
Zhang Jinsheng, Zhang Jiazhong, Wei Yingjie, Cao Wei
2010, 36(4): 411-414.
Abstract:
The relation between the impact loads on supercavitating underwater vehicle’s tail and the vehicle’s velocity was advanced firstly, with the method of combining the conclusions. Then, the structural response and deformation characteristics of the vehicle were studied by finite element method (FEM). The dominant frequencies of the acceleration responds were obtained. The maximum strains of the vehicle and their locations were gotten. And the relations between the dominant frequencies and the vehicle- velocities, the maximum strains of the vehicle and the vehicle- velocities were discussed. Simulation results indicate that there are two dominant frequencies. If the vehicle-s velocity is relatively small, the first dominant frequency should be chosen as reference frequency for structure design, and the second one should be chosen for relatively large velocity. The simulation results also indicate that the maximum deformations of the vehicle increase with the increase of the vehicle-s velocity. These results would be helpful for structure strength analysis and optimum structure design of supercavitating underwater vehicles.
Evaluation of XY-SAS model for separated flows
Gao Ruize, Xu Jinglei, Zhao Rui, Yan Chao
2010, 36(4): 415-419.
Abstract:
Xu-Yan-scale adaptive simulation (XY-SAS)is a new one-equation scale-adaptive simulation model proposed on the basis of average vorticity. It resolves the reynolds average navier-stokes(RANS)-large eddy simulation(LES) interface problem in detached eddy simulation (DES) with much smaller computation cost compared with pioneer SAS model. It was analyzed that XY-SAS model can provide proper damping of resolved turbulence at high wave number end of the spectrum compared with pioneer one. The limiter was removed based on mechanics analysis and this model was improved. Flat plate boundary layer and square cylinder were chosen as test cases for further evaluation of XY-SAS. The numerical results agree well with the experimental data, which proves that XY-SAS is an effective method for large scale separation flow and the model itself is complete. The model viscosity is determined adaptively according to the local flow behavior .A smooth transition is achieved from RANS to LES region.
Robust control method for disturbance rejection on axial magnetic bearing
Ding Li, Fang Jiancheng, Wei Tong, Zheng Shiqiang
2010, 36(4): 420-423.
Abstract:
Considering the requirements of robust stability and dynamic performance to reject disturbance, the problem of low bandwidth was analyzed when applying the disturbance observation(DOB) in the magnetic bearing. It was pointed out that the magnetic bearing was affected by outside low bandwidth noise, then a robust disturbance rejection method was proposed. The nominal model of the axial magnetic bearing was set up through dynamic analysis.And the DOB for internal loop controller was designed consisting of the inverse nominal model and a low pass filter with two relative rank. In external loop, a robust controller is developed using standard H analysis method after choosing proper weight functions to enhance robust stability. The simulation analysis and experiment result indicate that the method could extend the bandwidth of the DOB. Comparing to single controller, experimental result verifies that the robust disturbance rejection system controller has broad robust stability margin on parameter uncertainty and strong ability to eliminate disturbance.
Method and its system of Java source and byte code plagiarism detection
Li Hu, Liu Chao, Liu Nan, Li Xiaoli
2010, 36(4): 424-428.
Abstract:
A plagiarism detection approach to detect both Java source code and byte code was proposed. The proposed method compares Java source files or class files by multiple similarity measures developed to represent the syntax structures and semantic features of the programs. An efficient plagiarism detection tool using the proposed technique was developed to analyze plagiarism behavior of Java source code or class code. Statistical analysis and several graphical visualizations aid in the interpretation of analysis results. An experimental comparison with a typical commercial source code plagiarism detection tool as well as a case study by applying the tool to plagiarism detection with a set of manually modified programs were conducted. Experiment results show that the tool is more efficient and the proposed technique can recognize both exact copy and approximate copy, including those most of the types of source code transformations in program plagiarism behavior.
Robust facial expression recognition under occlusion condition
Xue Yuli, Mao Xia, Caleanu Catalin-Daniel, Lü Shanwei
2010, 36(4): 429-433.
Abstract:
A novel method for facial expression recognition which is robust to facial occlusion was proposed. Firstly, the face to be recognized is reconstructed using robust principal component analysis(RPCA), and saliency detection was used on the difference image of reconstructed face and the face to be recognized to obtain the facial occlusion region. Secondly, facial occlusion region of the face to be recognized was reconstructed by the reconstructed face using RPCA, and a novel reweighted AdaBoost classifier was used on the face after occlusion region reconstruction for facial expression recognition. Finally, facial expression recognition experiments were implemented in different occlusion conditions on Beihang university(BHU) facial expression database and Japanese female facial expression database and gained better recognition results than Adaboost method, showing that this method based on RPCA and AdaBoost is robust to kinds of facial occlusions.
Efficient arithmetic for space partitioning
Liu Zhongtie, Ma Dongli
2010, 36(4): 434-437.
Abstract:
An efficient algorithm was provided for space partitioning of triangular facets. The algorithm based on judging the position relationship between a 3D point and a unit cubic, and extended to the straight-lines and triangles- partitioning. A new method of triangle-facet targets fast partitioning was given. Its application in solving parametric curves and non-uniform rational B-spline (NURBS) surface was considered, and the solution to the situation of non-uniform partitioning was also given for consult. Compared with the binary space partitioning(BSP) trees technology, which was the most popular method, this arithmetic performed better than BSP in solving particular problem. The results prove the efficiency and reliability of the arithmetic.
Reusable method of requirement modeling
Tang Wenzhong, Deng Jingwen
2010, 36(4): 438-442.
Abstract:
A reusable requirement method of software (RRM) was proposed. The method defined a set of reusable requirement modeling process (RRMP) and combined with the software manufacture development of organization manager management domain. Requirement customing & application developing platform (RAP) was designed and came true. The mainline of RRMP covered domain analysis, design, application development based on component and other software development main process and set up several requirement model including domain model and application model. On the guide of RRMP requirement process, RAP for domain software automation development was developed. This platform based on the requirement, ran through the three important development stages of requirement analysis, software design and implement to a certain extent. It realized the automation development from requirement to implement. RAP platform includes three primary tools: graphic requirement customized tool, requirement development document produced tool and object application automatic development tool.
Environment referring factor based on proportional hazards model
Hong Dongpao, Ma Xiaobing, Zhao Yu, Zhang Lin
2010, 36(4): 443-446.
Abstract:
The relations of reliability and environment factors were described by the proportional hazards model (PHM). Using the model,the influences of environment factors on reliability were measured quantificationally. A calculation method of environment referring factor was proposed. In the method,the varied environment test data was used synthetically to improve the calculation precision of environment referring factor. The prediction of environment referring factor is becoming possible with the method. For complex environment,the generalized proportional hazards model was introduced to measure the interaction of varied environment on reliability. The model was established by radial basis function using the varied environment test data. The statistical inference of the environment referring factor for usual life distributions was put forward based on the PHM. The example shows that the method is reasonable and straightforward for engineering.
Sort-first rendering clusters with dynamic load balancing
Chang Hui, Lei Xiaoyong, Dai Shuling
2010, 36(4): 447-450.
Abstract:
To solve load imbalance problem among the nodes of rendering clusters, a sort-first rendering cluster was proposed. The deferred shading was used as major rendering method in the graphics applications of the proposed rendering cluster. Algorithms of deferred shading were classified as visibility testing, global pixel shading and local pixel shading. These algorithms were used to predict rendering load and converted to fragment shadings of rendering. Referring to the parallel computation performance of graphics processing unit (GPU), the sort-first rendering cluster predicted per-pixel rendering load within the whole screen space. On the base of the rendering load distribution, the proposed system subdivided rendering tasks and sorted them to each node of the cluster. According to the experiment results, the rendering load distribution can be predicted in real-time, the rendering tasks can be subdivided by the proposed sort-first rendering cluster. The higher and stable rendering performance of the system was obtained by the accurate load prediction.
Calculation methods for mean waiting time of maintenance support
Wen Jia, Kang Rui, Qiao Li, Cheng Hailong
2010, 36(4): 451-454.
Abstract:
In order to analyze and evaluate the timeliness of the support system during the design and development phase, the concept of waiting time of maintenance support (WTMS) was discussed. The impact factors of WTMS during the design process of the support system for the equipment was analyzed, including the frequency and location of the support activities, depth of maintenance of the support sites with different lines of maintenance, distribution of the support sites and allocation of the support resources in them, transportation means between different support sites and other factors.A calculation method for the mean waiting time of maintenance support was established based on statistics and matrix theory, which used the results of supportability analysis as input and considered the difference in calculation methods between the consuming support resources and occupying support resources. A case of the calculation method was provided to verify its feasibility and correctness.
Topology optimization design of frame structure for magnetic suspension control moment gyroscope
Liu Fengrui, Zhao Libin, Han Bangcheng, Fang Jiancheng
2010, 36(4): 455-458.
Abstract:
The finite element model of frame structure for magnetic suspension control moment gyroscope (MS-CMG) was established and a modal analysis was implemented. On the view of the numerical results the optimization opinions of the frame structure were presented. The topology optimization module for continuum structure was developed and plugged in ANSYS via its APDL and UIDL combined with the bi-directional evolutionary structural optimization method (BESO). Thus the dynamic topological optimization of frame structure was realized with the fundamental frequency as constraint condition and the minimum weight as the target function. In this procedure, a nominal stress was presented and a relationship between the nominal stress and the corresponding mode shape of frame structure was established, by which the BESO was extended to apply in field of dynamic optimization. At last, a frame structure configuration was provided with optimum layout of stiffness and mass was given. Comparing with the initial configuration, the optimum frame structure satisfied the design demands and the fundamental frequency increased by 11.49%, whereas the mass decreased by 5.65%.
Beyond visual range interception effectiveness evaluation of formation fighter
Zheng Jiang’an, Ma Dongli, Zhao Juanjuan
2010, 36(4): 459-463.
Abstract:
Evaluation problem of formation fighter’s beyond visual range interception effectiveness was studied by Lanchester theory. Comparing hypothesis condition of Lanchester theory with actual background of beyond visual range interception, detection and attack phases of interception mission were respectively described by secondary linear law and square law. On the principle of maximum commutative loss ratio, fire distribution was introduced into evaluation model. Dynamic models of detection effectiveness and attack effectiveness were established by using finding probability of airborne radar and hit probability of air-to-air missile respectively. A numerical example shows that value coefficient of assault side-s escort fighter significantly influences fire distribution of interception side, and interception side-s tactics selection attacks escort fighter first and then fighter bomber. Interception effectiveness could be improved both by using electronic countermeasure and enhancing fighter survivability, and the model is not only an easy calculation but also reasonable and effective.
Fatigue properties of S06 steel in long life regime under ultrasonic frequency
Hu Yanhui, Zhong Qunpeng, Zhang Zheng, Han Bangcheng
2010, 36(4): 464-468.
Abstract:
S06 steel is a new kind of precipitation-hardened stainless martensite steel. The fatigue properties between 106~109 cycles of S06 steel were tested with ultrasonic fatigue testing machine. Results show that fatigue failure takes place after 107 cycles and an endurance limit does not exist for S06 steel. Fatigue design on infinite life criterion can-t be used for components made of this steel. High-cycle fatigue testing method with the same loading mode and approximate specimen dimension with that of ultrasonic fatigue tests was used to study the fatigue properties between 104~107 cycles. The results show that fatigue strength of S06 steel is higher with ultrasonic fatigue testing method. Ultrasonic fatigue testing data was amended by comparing the fatigue strength coefficient of Basquin formula. The P-S-N curves at 106~109 cycles of S06 steel considering cracking initiation mechanism and frequency effect were fit with linear variance regression analysis method and the curves can be used for fatigue design of engineering components with long lifetime.
Numerical simulation of combustion for GH2/GO2 gas-fired shear coaxial injector
Li Mao, Du Zhenggang, Jin Ping, Cai Guobiao
2010, 36(4): 469-473.
Abstract:
To research the performance of the gas-fired injector in full flow staged combustion(FFSC) cycle engine, a gas-fired shear coaxial injector was simulated to obtain the performance of combustion. The gaseous hydrogen and gaseous oxygen (GH2/GO2) preburners were utilized to provide the hydrogen-rich gas at the temperature of 758 K and the oxygen-rich gas at the temperature of 676 K. It was performed to analyze the influence of the oxygen-rich gas pressure drop, the ratio of hydrogen-rich gas velocity to oxygen-rich gas velocity, the thickness of oxygen-rich gas injector and the recess of oxygen-rich gas injector on combustion performance. All the operating condition have the same structure of chamber, mass flux, gas-fired temperature. Influence of the four parameters were obtained. The results provide reliable criterion for the design of gas-gas shear coaxial injector.
Unmanned aircraft vehicle path planning based on image skeleton and greedy algorithm
Zhao Yongjia, Dai Shuling
2010, 36(4): 474-477.
Abstract:
To ensure the mission success rate for low attitude penetration, a path to meet the restriction of the unmanned aerial vehicle-s(UAV) capability with high survivability must be planned.By improved the traditional artificial potential field approach, a novel method was presented based on image skeleton and the greedy algorithm generating optimal UAV flight paths over hostile territory. A weighted graph was generated by skeletonizing the permitted area and calculating flight cost of the skeleton branches based on a threat model. The rough estimate of flight path was given by searching in the graph using Dijkstra algorithm. The flight path was dynamically estimated through minimization of an objective function. The minimization was done efficiently using an improved greedy algorithm with the curvature controllable for local optimization to meet the requirements of the minimum turning radius and the shortest range. The experimental results demonstrate the potential of the proposed method.
Error modeling and compensation for indirect stabilization electro-optical targeting system
Xu Yefeng, Meng Qingji, Wang Lei, Feng Peide
2010, 36(4): 478-481.
Abstract:
In order to improve the precision of indirect stabilization electro-optical targeting system, the line-of-sight error model of the system was established. The rotation-modulation approach was studied for alleviating the drift error of the gyro. The speed measurement method by combining tachogenerator and photoelectric encoder was proposed for decreasing quantization noise of the photoelectric encoder. The relationship between the installation error of gyro & camera and line-of-sight stabilization error was analyzed, and the installation error was effectively estimated by using electro-optical tracking method. The theoretical analysis is consistent with the real experimental results. The simulation results show when the system error was completely compensated, the accuracy of indirect stabilization system based on low performance micro electronic mechanical system (MEMS) gyros was equal to the direct stabilization system based on medium performance fiber gyros.
Design and parametric analysis of front suspension of new sport car
Jiang Libiao, Liu Yonghua, Gu Fangde, Wei Tao
2010, 36(4): 482-485.
Abstract:
To satisfy people-s pursuing for individuality, an assumption of designing a new sport car was presented. Front double-wishbone suspension was designed basing on the whole vehicle-s design parameters and suspension design theories. Suspension-s virtual prototype modeling was built in ADAMS/View. By doing the design study, the assembly angles with horizontal plane of lower control arm and upper control arm were the best influential design variable on each appraisable target of suspension-s performance. These two design variable could change not only the size variation but also the trends. Cut-off point of tie rod influences toe angle greatly impacted wheels toe angle. When consider performance change and variation trend influences, optimization design was done. Lack of cornering and back to performance were ensured. The optimization results meet the demands of suspension design, and show that the suspension performance parameters achieve the ideal variation range, and is the combination of production, experiment and research.
Prediction modeling based on Bayes support vector machine
Hu Wenliang, Wang Huiwen
2010, 36(4): 486-489.
Abstract:
To solve the uncertainty, nonlinear and coupling problem of statistical data, Bayes support vector machine(BSVM) was proposed to predict their development trend. Herein, the uncertainty of data was described as BSVM weights with Gauss distribution. Based on the prior probability and Bayes theory, the parameters evaluation of BSVM was transformed into parameters optimization of posterior distibution, which can be obtained by the prior probability and Bayes theory. The nonzero vector as correlative support vector machine was selected, and the multiple dimension prediction model based on time serials and its parameters distribution were established. Considering the input of BSVM as random variable during every iterative process, the output of BSVM can be obtained with uncertainty transferring. Since BSVM can describe the influence of random variables and its tranferring, it can overcome the uncertainty and dependence influence and the prediction results approach to the real condition. Application indicates that the prediction of high-tech enterprise development based on BSVM can approach the actual condition with high precision and robust.
Expert system on cost control for complex equipment
Wang Fei, Wei Fajie
2010, 36(4): 490-494.
Abstract:
Because the research and development of complex equipment have a long cycle, cost too much and are high-risk,a cost control structure of large scaled complex equipment development according to the feed-forward control was presented. At the same time the method of knowledge representation and inference engine were given based on this structure,which enabled the knowledge base and inference engine to be realized independently. This system could make real-time monitor and prediction of manufacturing cost with the application of cost control system and knowledge management system. For the input problem, the system gave decision support to administrator after reasoning, which offered a feasible basis for the realization of intelligent management on research cost. Because the realization of knowledge base and inference engine was relatively independent, this system had better maintainability and reliability compared to conventional expert system.
Spectral interpolation measurement algorithm of high accuracy on signals in power network
Wu Jing, Jin Haibin
2010, 36(4): 495-499.
Abstract:
In the condition of asynchronous sampling, the long-range spectral leakages, short-range spectral leakages and negative frequency spectral leakages will be caused by using Fourier transforms to analyze the harmonics in power network. To improve the measurement accuracy, the spectrum leakage error expressions were derived. For power frequency alternating current signals, a new error correction algorithm based on interpolation was presented to eliminate both of the short-range spectral leakage errors and the negative frequency spectral leakage errors at the same time. Compared with the previous interpolation algorithms, the simulations and practical measurements show the new algorithm need not weight the samples by special windows and can measure out the accurate frequency, amplitude and phase of the signals even if the asynchronous sampling is very serious. It is applicable for the precise measurement fields requiring a higher accuracy.
Query clustering using user-query logs
Jia Rongfei, Jin Maozhong, Wang Xiaobo
2010, 36(4): 500-503.
Abstract:
A new query clustering method on user-query log was presented. Traditional clustering techniques focused on queries and click-through logs, which are often sparse. The average cluster size is often small. In contrast, the user-query log is much denser as well as noisier. To reduce the influence of the noises and discover similar queries, queries visited by the same user at the same session were assumed to be mostly similar. Based on the assumption, a new similarity measure using query co-occurrence relations was calculated to create query neighbor vector space. The queries were represented by vectors consisting of their neighbors. The similarity function for clustering was calculated based on the query neighbor vectors. An adjusted clustering method of density-based spatial clustering of applications with noise(DBSCAN) was applied to generate the clusters. Experiments on a real dataset of 95262 queries show that 79.77% precision and 48.21% recall is achieved and the average cluster size achieves 51.