最新刊期

    50 6 2016
    • Vol. 50, Issue 6, Pages: 1-8(2016) DOI: 10.7652/xjtuxb201606001
      摘要:An expression driven brain-computer interface system is proposed to solve the low recognition rate and huge individual differences of spontaneous electroencephalogram(EEG)signal, and simulation and experimental verification are conducted. A neural mass model is used in mechanism modeling and simulation of expressions driven EEG signal, and the results show the related brain areas of facial expression and the frequency distribution characteristics of expression driven EEG signals. A classification and recognition method based on the wavelet transform and an artificial neural network is put forward, to improve the recognition rate of expression driven EEG signal. This research elaborates that it is the prefrontal cortex and the limbic system to cooperate on the generation and control of expression driven EEG signals, and verifies the feasibility of the classification and recognition method whose highest off recognition rate is 85%.  
      关键词:expression control;electroencephalogram;mechanism modeling;wavelet transform;artificial neural network   
      1511
      |
      4
      |
      9
      <HTML>
      <L-PDF><Meta-XML>
      <引用本文> <批量引用> 73111276 false
      更新时间:2025-07-09
    • Vol. 50, Issue 6, Pages: 9-14+82(2016) DOI: 10.7652/xjtuxb201606002
      摘要:Two-agent single-machine scheduling problems with a deterioration job of which the actual processing time is defined as a function of its starting time are investigated. Optimization algorithms of polynomial time under some certain constraints are proposed. The optimization target of the problem is formulated as to minimize the total weighted completion time or total weighted discounted completion time of one agent while ensuring that all the jobs on the other agents are not delayed. This optimization problem is proved NP-hard and the structural properties of its optimal solutions under general and special conditions are illustrated. Optimization algorithms of polynomial time for solving these scheduling problems are proposed under some constraints. First, the jobs from two agents are respectively sequenced according to their optimal structural properties. Then, they are combined following the strategies that ensure the optimization target of two agents. Simulation results and comparisons with scheduling results of the simulated annealing(SA)and the branch-and-bound algorithms show that the CPU-times of the proposed algorithms are similar with the SA's that has an error rate greater than 0.3% with the optimal solutions, and much smaller than that of the branch-and-bound algorithm.  
      关键词:scheduling;single-machine;two agents;deteriorating jobs   
      655
      |
      4
      |
      0
      <HTML>
      <L-PDF><Meta-XML>
      <引用本文> <批量引用> 73107015 false
      更新时间:2025-07-09
    • Vol. 50, Issue 6, Pages: 15-22(2016) DOI: 10.7652/xjtuxb201606003
      摘要:The traditional nonlinear blind source separation(NBSS)algorithms often fall across the problem of local optimal solution to lead a lower separation precision. An NBSS algorithm based on improved particle swarm optimization(PSO)is proposed, where the multilayer perception(MLP)is used to fit the inverse of the nonlinear mixed process, and the mutual information between separated signals is regarded as the optimization objective(Fitness function of PSO)to realize the optimization of parameters in MLP. However, the canonical PSO algorithms usually suffer from particle premature problems and are easy to get into local optimal solution. Thus crossover and mutation operations are applied to the particles with lower fitness according to probability mechanism to efficiently increase the diversity of the particles, and the premature problem of canonical PSO is solved. The simulations and experiments show that compared with the linear blind source separation algorithm and the NBSS algorithm based on canonical PSO, the proposed algorithm enables to extract pure independent source information from mechanical information with nonlinear mixing and improve the separation precision of nonlinear mixed signals.  
      关键词:nonlinear blind source separation;particle swarm optimization;particle premature;crossover and mutation   
      721
      |
      4
      |
      1
      <HTML>
      <L-PDF><Meta-XML>
      <引用本文> <批量引用> 73114226 false
      更新时间:2025-07-09
    • Vol. 50, Issue 6, Pages: 23-29(2016) DOI: 10.7652/xjtuxb201606004
      摘要:A background subtraction method based on decomposition of low-rank and structured sparsity matrices is proposed to solve the problem that detection results are sensitive to noise and incomplete caused by ignoring the relationship between foreground pixels in traditional foreground detection methods based on matrix decomposition. The method takes the structural distribution of the foreground into account, and a structured sparsity constraint is used on the foreground pixels. Moreover, a two-stage framework based on motion saliency is introduced to address the parameter setting issue in dynamic background videos and to tune regularization parameters adaptively. Motion block candidates are obtained by using the low rank and structured sparsity decomposition in the first step. Then, motion saliency analysis is applied to these candidates and the adapt block sparsity decomposition is used to detect the foreground in the second step. Experimental results show that the performance of the proposed method is more adaptive than the existing foreground detection methods based on matrix decomposition in complex videos, and that the proposed approach outperforms the state-of-the-art methods according to the precision and recall results on dataset I2R.  
      关键词:foreground detection;background subtraction;matrix decomposition;low rank representation;structured sparsity   
      871
      |
      4
      |
      2
      <HTML>
      <L-PDF><Meta-XML>
      <引用本文> <批量引用> 73110154 false
      更新时间:2025-07-09
    • Vol. 50, Issue 6, Pages: 30-35(2016) DOI: 10.7652/xjtuxb201606005
      摘要:A resource allocation method for clusters of streaming media servers in multi-version VoD is proposed to solve the problem of the inappropriate resources allocation for clusters of streaming media servers in large-scale mobile learning(m-learning)system. The method analyzes user's historical learning logs and mines user's behavior characteristics, and then it utilizes the queueing theory to establish a resource allocation model for the cluster of streaming media servers. The method uses the model to predict the user arrival rate in real-time and then allocates appropriate resources dynamically. Simulation results show the correctness and efficiency of the proposed method. The average service rejection rate and average resource utilization rate are around 1% and 80% respectively. The proposed method can ensure the user experience satisfaction and reduce service cost.  
      关键词:m-learning;multi-version VoD;user behavior analysis;queueing theory;resource allocation   
      744
      |
      4
      |
      0
      <HTML>
      <L-PDF><Meta-XML>
      <引用本文> <批量引用> 73107126 false
      更新时间:2025-07-09
    • Vol. 50, Issue 6, Pages: 36-41(2016) DOI: 10.7652/xjtuxb201606006
      摘要:A neural network model for the spool throttling groove topology combination of multi-way valves is proposed to deal with the problem that there exist various combinations of topological form expression of throttling groove structure on spool and huge degrees of freedom of analysis of fluid-structure coupling response in the design of multi-way valves. The throttling groove topological structures are classified into semicircle groove, U-shaped groove, round hole groove and so on, and form different parameterized combinations of these structures based on validation of analytical results in using the multi-way valve dynamic characteristic test. The orthogonal experiment method is used to obtain three-dimensional fluid analytical responses of the multi-way valve for each parameter combination, then these responses are used as training samples of a back propagation neural network to achieve the neural network expression of the throttling groove topology combination. The initial weights and threshold of training process are optimized by an evolutionary neural network to improve prediction accuracy of the model, so that accurate predictions for non-training samples are achieved. The results show that the proposed model provides a new idea for topological form expression in the design of spool throttling groove structures, and has practical significance for digital design of multi-way valves.  
      关键词:multi-way valve;throttling groove;topology;neural network;digital design   
      823
      |
      4
      |
      1
      <HTML>
      <L-PDF><Meta-XML>
      <引用本文> <批量引用> 73105771 false
      更新时间:2025-07-09
    • Vol. 50, Issue 6, Pages: 61-67(2016) DOI: 10.7652/xjtuxb201606010
      摘要:To estimate the continuous movement of human lower limb during walking, a regression model which relates the surface electromyography(EMG)and the movement variables of the lower limb joints is constructed. The joint movement angles of lower limb are calculated accurately based on optical motion capture system, then the surface EMG signals are sampled from the main muscles directly concerned with the lower limb motion; the muscle activities are extracted, and a deep auto-encoder(DAE)network with restricted Boltzmann machines(RBM)is realized, by which the multi-channel processed surface EMG signals are encoded in low dimensional space and the optimal features are extracted. The nonlinear model mapping the EMG features to sagittal surface movement angles is established with back propagation(BP)neural network. Extensive experiments indicate that the features extracted with the deep auto-encoder(DAE)network are outperformed principal components analysis(PCA); the movement angles of lower limb joints can be estimated continuously and precisely with the regression models and the mean square error(MSE)between the estimated values and real values is reduced by 25%-35% compared with the traditional method. The proposed strategy is expected to develop human-machine interaction interface technology for the achievement of continuous bioelectric control and the improvement of motion stability between human and machine, especially for lower limb wearable intelligent equipment.  
      关键词:surface electromyography;continuous motion recognition;restricted Boltzmann machines;deep auto-encoder;back propagation network   
      664
      |
      7
      |
      5
      <HTML>
      <L-PDF><Meta-XML>
      <引用本文> <批量引用> 73112691 false
      更新时间:2025-07-09
    • Spectrum Sensing Method of Narrow-Bands Based on Peak Feature of Spectrums

      Vol. 50, Issue 6, Pages: 68-74(2016) DOI: 10.7652/xjtuxb201606011
      摘要:A spectrum sensing method based on local density of spectrum peaks(MTSE-MLD method)is proposed to address the problem that the spectrum sensing of narrow-bands under the blind condition is not ideal in low SNR performance. Power spectrums of received signals are estimated by the multitaper spectral estimation(MTSE), and their maximum spectrum peak values are obtained to form a sequence of spectrum peaks. It follows from the idea of clustering that the maximum local density(MLD)of the spectrum peak is used as the detection statistic. Theoretical derivation and simulations verify that the statistic obeys an extreme value distribution of type II. A detection threshold with a given false alarm probability is obtained, and the blind detection of signals is realized. The MTSE-MLD method does not need any priori information, and is not sensitive to any noise variance. Experimental results show that the method is suitable for the detection of various common signals under the condition of additive Gauss white noise channel and multi-path fading channel. A comparison with the method based on eigenvalue energy under the same amount of data shows that the detection performance of the proposed method improves by 1 dB.  
      关键词:spectrum sensing;clustering;extreme value distribution;multitaper spectral estimation   
      767
      |
      4
      |
      0
      <HTML>
      <L-PDF><Meta-XML>
      <引用本文> <批量引用> 73109825 false
      更新时间:2025-07-09
    • Learning New Words for Pronunciation Lexicon from Audio Data

      Vol. 50, Issue 6, Pages: 75-82(2016) DOI: 10.7652/xjtuxb201606012
      摘要:A self-learning method of new pronunciation lexicons based on a hybrid speech recognition system is proposed to solve the problem that the existing self-expanding methods of pronunciation lexicons can only learn new words from text data but cannot learn from audio data. The method utilizes both the syllables and the graphones hybrid systems to recognize the out-of-vocabulary words in the audio data and then obtains as many new words with their pronunciations as possible by using the complementary information of the two systems. Then the new word and its pronunciation candidates are optimized using a perceptron model and a maximum entropy model to reduce the error rate. Finally, the lexicon is expanded and the language model parameters are updated by using syntactic and semantic information. Experimental results of continuous speech recognition on Wall Street Journal speech database show that the proposed method learns new words from audio data effectively, and the accuracy is greatly improved by using the data optimization strategies. The extended lexicon system yields a relative gain of 13.4% over the base line system in terms of word error rates.  
      关键词:speech recognition;pronunciation lexicon;new words learning;out-of-vocabulary words   
      635
      |
      4
      |
      0
      <HTML>
      <L-PDF><Meta-XML>
      <引用本文> <批量引用> 73112839 false
      更新时间:2025-07-09
    • Surface Error Measurement System for Bearing Ball with Laser Interferometry

      Vol. 50, Issue 6, Pages: 83-89(2016) DOI: 10.7652/xjtuxb201606013
      摘要:To avoid the influence from sphericity error of spherical mirror on the results in bearing ball measurement with laser interferometry, a Tyman-Green interference method by a plane reference mirror instead of a spherical mirror for the spherical surface error measurement of bearings ball is proposed and a prototype interferometer is developed. Following phase shifting interferometry principle, the interferometer system controls piezoelectric ceramic transducer phase shifter to move the plane reference mirror to obtain more interference images. The images are captured by charge-coupled device camera and then the spherical surface error is evaluated by phase unwrapping. Zernike polynomials fitting is chosen to eliminate the alignment errors from out-off-focus. Experiments verify its effectiveness and robustness, and the root mean square difference of results between surface error measurement system and ZYGO interferometer system is 0.004λ while the difference of peak-to-peak value is 0.001 6λ between the two systems. The method can be expected to apply to the other fields of roundness measurement.  
      关键词:phase-shift interferometry;bearing ball;wavefront fitting;alignment error   
      780
      |
      5
      |
      1
      <HTML>
      <L-PDF><Meta-XML>
      <引用本文> <批量引用> 73112547 false
      更新时间:2025-07-09
    • Vol. 50, Issue 6, Pages: 90-96+158(2016) DOI: 10.7652/xjtuxb201606014
      摘要:A novel method for machine tool multi-body system restrained by stiffness field is proposed to solve the difficulty of discontinuous machining error prediction result. According to the characteristics of multi-body system kinematics solution, the machine tool component is considered as an elastic body, and the lower body array and transfer matrix are used to describe the motion relationship between the machine tool components. Following the condensation approach of stiffness matrix and the least square evaluation, the stiffness model under moving load of machine tool component can be obtained. Then Jacobian matrix is taken to construct the machining error prediction model. This method is verified in a boring milling machining error prediction for an engine cylinder head. The combined machining error prediction method enables to optimize the angle of clamping in engine shell machining process. Describing multi-body system by lower body array facilitates evaluating errors in united machining process. According to the combined machining error prediction, the machining precision can be increased by 35% at working degree of 90°.  
      关键词:numerical control machine tool;multi-system theory;error prediction;manufacturing space;process optimization   
      700
      |
      5
      |
      1
      <HTML>
      <L-PDF><Meta-XML>
      <引用本文> <批量引用> 73112396 false
      更新时间:2025-07-09
    • Vol. 50, Issue 6, Pages: 104-109(2016) DOI: 10.7652/xjtuxb201606016
      摘要:An improved sliding mode observer(SMO)with the feature of identifying stator parameters in real time for permanent magnet synchronous motor(PMSM)control is proposed to address the case that the stator resistance and the inductance randomly change when a PMSM is running. An approximate saturation function is used as the switch function in the improved SMO control system to reduce the system chattering. The improved SMO estimates the stator resistance and inductance values by employing Lyapunov stability theorem on the premise of ensuring the stability of the observer and feedbacks them to the PMSM model to improve the estimation accuracy of rotor's position and velocity. The rotor's position and velocity information are estimated from the back electromotive force based on the principle of phase-locked loop. Simulation results show that the improved SMO quickly and accurately tracks the position and the velocity of a rotor, and has an attractive dynamic performance by estimating the resistance and inductance of the stator in real time and correcting the PMSM model.  
      关键词:permanent magnet synchronous motor;sliding mode control;sensorless control;approximate saturation function;parameters estimation   
      909
      |
      4
      |
      13
      <HTML>
      <L-PDF><Meta-XML>
      <引用本文> <批量引用> 73112746 false
      更新时间:2025-07-09
    • Vol. 50, Issue 6, Pages: 110-115(2016) DOI: 10.7652/xjtuxb201606017
      摘要:A method that combines the finite element method and Newton-Raphson iterative method is applied to study the effects of bit pattern media(BPM)on the static characteristics of slider in the hard disk drive. The static characteristics of sliders with different structures and geometry dimensions of BPM such as pitch angle, roll angle, air bearing film thickness, pressure and load capacity are calculated. The numerical results show that the bit shape and its dimension parameters have significant effects on the static performances of the slider. Compared with the smooth disk, when the slider is flying over the bit pattern media, the pressure distribution in the central region which is close to its read/write head shows a discretization feature. The gas-film load capacity and minimum film thickness of the slider gradually decrease as the bit height increases, while the changes of other characteristics are closely related with the bit distribution. Moreover, a comparison with a cylindrical bit pattern media shows that the smaller roll angle may produce lager maximum film pressure, load capacity and minimum film thickness, and that the static characteristics of the slider flying over cuboidal bit pattern media are better than those over cylindrical bit pattern media. These results are of great significance for the design of structure and geometry parameters of bit pattern media.  
      关键词:hard disk drive;bit pattern media;gas lubrication;static characteristics   
      703
      |
      4
      |
      1
      <HTML>
      <L-PDF><Meta-XML>
      <引用本文> <批量引用> 73105727 false
      更新时间:2025-07-09
    • Vol. 50, Issue 6, Pages: 116-121(2016) DOI: 10.7652/xjtuxb201606018
      摘要:A novel optimal torque control allocation strategy for the control of the power system that consists of the DC motor/generator and the diesel engine is proposed to improve the energy efficiency and to decrease the torque amplitude of variation for a parallel hybrid power system with an energy regenerator by capacitor for hydraulic excavator. Two work areas with different rotate speeds are built based on load forces. Curve fitting is used to model the character between the torque and the energy efficiency, and a penalty function is built to decrease torque range, then multi-object optimal cost functions are put forward. Experiments are performed by simulation program on a simulated experiment platform. The results and a comparison with the constant torque strategy show that the proposed strategy decreases the torque amplitude of variation by 37.9%, the whole energy by 8.74%, and the fuel consumption by 8.9%. It can be concluded that the new strategy improves the energy consumption, the fuel consumption and the impact on mechanics system caused by torque range.  
      关键词:energy efficiency;parallel hybrid system;hydraulic excavator;control strategy   
      895
      |
      4
      |
      4
      <HTML>
      <L-PDF><Meta-XML>
      <引用本文> <批量引用> 73105841 false
      更新时间:2025-07-09
    • Multi-Point Location Method of Large-Scale Thin-Wall Component

      Vol. 50, Issue 6, Pages: 122-129(2016) DOI: 10.7652/xjtuxb201606019
      摘要:To analyze the relationship between workpiece machining deformation and multi-point location/support distribution, a new optimization method is proposed to amend multi-point location/support distribution. Depending on the idea of representing the sensitivities of machining deformation with finite difference method, the analytic expressions for sensitivities of machining deformation with respect to location/support distribution are established. The location/support distribution is adjusted along the most sensitive direction to restrain machining deformation and improve process system stiffness. A multi-point location of large-scale thin-wall component is taken as an example, and the searched initial distribution of location/support point is regarded as the initial value for the global optimization of multi-point location/support distribution. A comparison with non-optimized workpiece machining path deformation shows the maximum machining deformation is reduced by 34.5%, which is limited below 0.669 4 mm far less than the non-optimized value of 1.022 mm.  
      关键词:thin-wall component;multi-point location;distribution optimization   
      821
      |
      4
      |
      2
      <HTML>
      <L-PDF><Meta-XML>
      <引用本文> <批量引用> 73112919 false
      更新时间:2025-07-09
    • Vol. 50, Issue 6, Pages: 130-135(2016) DOI: 10.7652/xjtuxb201606020
      摘要:An experiment platform is established to measure lightning impulse characteristics of carbon fiber reinforced polymer(CFRP), on which the experimental circuit and structures of CFRP are developed to propose a suitable two-electrode testing method. The dynamic characteristics of CFRP under non-destructive lightning current A wave(TP=30.2 μs, Td=72.2 μs)are investigated. The experimental results show that the carbon fiber composite material has a strong conduction capability under lightning impulse, and its impact resistance exhibits significantly nonlinearity, and a close relationship between surface coating materials and internal layer-structures is revealed. Considering the internal ply structure, surface material and conductivity of carbon fibers, micro carbon nanotubes layers and polymer, the lightning conducting model of CFRP is established. The inductance characteristics, nonlinear dynamic resistance characteristics as well as the linear trend in dynamic V-I characteristics can be well explained.  
      关键词:carbon fiber reinforced polymer;non-destructive impulse testing methods;dynamic impedance characteristics;non-linear characteristics;inductance characteristics   
      717
      |
      4
      |
      7
      <HTML>
      <L-PDF><Meta-XML>
      <引用本文> <批量引用> 73106965 false
      更新时间:2025-07-09
    • Cold Energy Recycle from Cryogenic Wind Tunnel Exhaust System

      Vol. 50, Issue 6, Pages: 136-142(2016) DOI: 10.7652/xjtuxb201606021
      摘要:To promote the usage of potential cold energy from exhaust system of cryogenic wind tunnel, realize scientific construction and energy saving, the status and methods of cold energy recycle are discussed. The operating processes and features of exhaust system in high Reynolds number testing are introduced and the cold energy from cryogenic wind tunnel exhaust system is calculated preliminarily. Based on the process analysis of oversea cold energy recycle schemes from cryogenic fluids gasification and the technical features of domestic cold energy recycle schemes, the recovery and efficient utilization of cold energy from liquid nitrogen spray cooling in cryogenic wind tunnel are studied. Considering the technology feasibility and economical efficiency, three expected cryogenic energy recycle schemes are provided, including a storage of the exhaust energy for civil air-conditioning, re-liquefying the gaseous nitrogen through an air separation process and using solid nitrogen as the cooling source. The advantages and disadvantages of the above-mentioned three approaches are analyzed. The first scheme has a low recycling rate because of the necessary additional cold-consuming equipment; the second scheme has an acceptable recycling rate and a good economical efficiency without additional investment; the third scheme has a good recycling rate but the further higher investment. Thus an appropriate recycle scheme and a reasonable wind tunnel testing process are important for cryogenic wind tunnel exhaust energy recycling.  
      关键词:cryogenic wind tunnel;liquid nitrogen;cold energy recycle   
      756
      |
      4
      |
      9
      <HTML>
      <L-PDF><Meta-XML>
      <引用本文> <批量引用> 73111534 false
      更新时间:2025-07-09
    0