In this article, the demand of urban bus operation condition on vehicle power system is analyzed from an energy perspective. Based on the analysis, for Hybrid Electric Vehicle (HEV) with composite electric power supply system, a new method is proposed to assign the required power to each energy source. Using this method, an optimal matching of three energy sources, the motor, battery and super capacitor, is conducted, in which the fuel economy, energy conversion efficiency, the service life and cost of the electric power supply system are taken as the reference indicators. The simulation on redeveloped ADVISOR shows that the matched scheme can achieve the desired performance.

This paper describes a three-step iteration method for determining the best separation factor of power-split hybrid electric vehicle based on system instantaneous efficiency. The function of the relationship between the best separation factor and the vehicle speed and power demand is obtained. Then the instantaneous optimal control algorithm of power-split hybrid electric vehicles is developed based on the best separation factor. As compared to the control algorithm based on the engine optimal curve, economical efficiency of the instantaneous optimal control algorithm based on optimal separation factor of system is improved significantly.

To improve the accuracy of insulation detection between the high-power battery system and the chassis of electric vehicles under complicated working conditions, this paper builds a new insulation detection model. This model takes the battery internal resistances into consideration for insulation detection. Based on the proposed model, a Reliability Algorithm (RA) is developed. This algorithm realizes the reliability measurement of the detection dataset according to the change interval of the total voltage, and then selects a binary dataset that has the maximum reliability to calculate the system insulation resistance of the battery system. In order to restrain the random measurement noise, the moving average filter is applied to process the results. The experiments carried out the test platform of electric vehicles and the numerical simulation show that the internal resistance model and the reliability algorithm for insulation detection have good estimation accuracy and high robustness.

In order to accurately characterize the dynamic performance of vehicles, a simplified model of commercial vehicles is established and the key parameters of this model are identified by means of the combination of least square algorithm and genetic algorithm and using Trucksim data. Simulation of the dynamic working condition of the model is carried out using the MAP of the identified key parameters. The simulation result is compared with the Trucksim data, which indicates that the MAP of the identified key parameters can satisfy the need to characterize the actual state of vehicle using the simplified model and lays a foundation for vehicle stability control.

In this paper the overall structure of Electromechanical Brake (EMB) is introduced;and the Support Vector Regression (SVR) is employed in the fault detection and diagnosis of the EMB sensors. First, the fault prediction model of the EMB sensors is built using SVR algorithm. Then, the residual sequences are generated from the redundancy information of the current sensor, force sensor and rotate speed sensor. Finally, the method is verified by experiments. Experiment results show that the proposed SVR algorithm can effectively detect the fault of the EMB sensor system that does consider the precise model of the system. The algorithm is applicable to complicated EMB systems.

In order to enable the intelligent vehicle to travel following planned trajectory smoothly under different road conditions, a new cognitive architecture Adaptive Control of Thought-Rational (ACT-R) based dynamic optimal trajectory planning method was introduced. The new method combines the mathematical tool for optimal control problem and the ACT-R model. The ACT-R model is the core of the new method and the trajectory is generated by the optimal control method. The multi-weights of the cost function are adjusted dynamically by intelligent optimization and the optimized trajectory is generated. The simulation results of obstacle avoidance show that the method is effective and feasible, and the optimal reference trajectory can be generated dynamically.

The paper gives dynamic equations of climbing over obstacles, judging standard of climbing over obstacles capability was proposed based on dynamic equations. State motion of vehicle is analyzed based on climbing over obstacles by single front wheel, driving torque of climbing obstacle was obtained, and its influential factors gravity allocation was analyzed. Finally, the simulation analysis gives conclusion the theory inference result is correct.

This study focuses on a single line transit scheduling problem, considering the characteristics of the unbalanced two-way and one-way distribution of passenger demand, a bi-level programming model of transit mixed scheduling and procurement is proposed. The relationship between the mixed scheduling and its procurement scheme was analyzed from the viewpoint of system theory. The upper-level model takes the minimal bus purchase expense and operation cost as the targets, and the required standard of service as the constraint. The lower-level model simultaneously optimizes the frequencies of inter-zone buses and two-way whole-line buses as well as the turning location of inter-zone buses in order to minimize the passenger travel time costs. For comparison, the model for single scheduling and procurement was also studied. Furthermore, numerical example and demand sensitivity analysis were given to verify the effectiveness and advantage of the proposed model. In addition, data comparison and analysis show that the mixed scheduling can save the transport capacity, passenger travel time and the operation cost under unbalanced demand situations. The more severe the unbalanced demand, the more superior the proposed model with the same bus procurement scheme.

To reduce vehicle delay at recurrent bottlenecks on freeways, this study aims to develop a control strategy of Variable Speed Limits (VSL) to improve the traffic efficiency near bottlenecks. First, the relationship between the capacity drop and traffic efficiency was analyzed. Then a control strategy of VSL that aims to prevent the capacity drop was proposed. The implementation of VSL at several different bottlenecks and the expected effects were discussed. The proposed strategy was evaluated for a merging bottleneck via simulation. The results show that the total travel time was reduced by 25.5% by the VSL, and the total delay was reduced by 56.1%. The traffic efficiency in the vicinity of recurrent bottleneck was improved by this VSL control strategy.

In order to solve the problem of large amount of data encountered in solving the shortest path of urban road network, in this paper, an ant colony optimization algorithm based on cloud computing is proposed. The algorithm combines the simulated annealing algorithm, to compensate for the shortcomings of the ant colony algorithm. Compared to Message Passing Interface (MPI) parallel ant colony algorithm, with the increase in the nodes the running speed of the proposed algorithm is accelerated noticeably.

To analyze the effects of external operation environment on the performance of bus routes, a model framework is developed for independent measurement of operation efficiency and quality efficiency of bus routes. The model adopts a more sophisticated two-stage approach based on Data Envelopment Analysis (DEA) and Stochastic Frontier Analysis (SFA). Furthermore, four operation environment factors are introduced at the microscopic level of bus routes, which reflect the effects on the performance, and a set of performance indicators is established. Results of empirical analysis show that, the developed model framework is rational;the operation environment indicators significantly influence the operation efficiency and quality efficiency, especially the later. With respect to the operation efficiency and quality efficiency, both the average intersection density and average dwell time per bus stop are unfavorable;high percentage of priority lanes and high average pure operation speed are favorable;long route length and high bus stop density are helpful for operation efficiency but unhelpful for quality efficiency.

The evolvement process of hierarchies of simplified urban road networks with different areas and land utilization structures was investigated by computer simulation. The self-organization system framework of the hierarchies of urban roads was established. The difference of total traveling time and number of accidents before and after the updating of the hierarchies were regarded as the guide line that whether the road hierarchies should be updated. Matlab software was used to update the road hierarchies until the network achieving stabilization. Six road networks with different areas were studied to reveal the self-organization phenomenon of road network. It was found that, in the circumference of high hierarchy roads, other land-use types take a small proportion;in the circumference of road networks with the same area, different land-use types have consistent distribution.

A pedestrian detection method by radar and vision information fusion is proposed. First,a coordinate transformation model of two sensors is established,by which depth information of radar scanning can be mapped to image.Then,radar scanning data are divided into a number of point clusters.Considering the physical attribute of pedestrian, the radar point clusters are selected again to obtain effective point cluster information.According to camera imaging rule, the relationship between pedestrian imaging region and his/her position is deduced.As a result, the Region of Interest (ROI) in pedestrian detection is determined.On this basis, the Histogram of Oriented Gradient (HOG) feature of the ROI is extracted,and the hypothesized pedestrian is verified using Support Vector Machine (SVM) classifier. Experiment results indicate that the proposed method is able to detect pedestrian in real time,and the accuracy rate can exceed 93%.

In this article the game theory is employed to describe the complicated game relationships between the parking administrators and the parkers, and among the parkers in Central Business District (CBD). A parking pricing model is developed using two-level game theory, and Stackelberg theory is used to seek the solution of the model. With the parking pricing model, the optimal parking rates for inside/outside CBD and peak/off-peak hours are determined. Research results indicate that the developed model can be used to find the optimal parking rates for urban CBD based on consideration of both time and special factors. It can be also used to balance the interests between the administrators and the parkers. The results can provide valuable information on how to develop an effective parking pricing program to solve the parking problem in the urban CBD.

The paper employed the orthogonality of rotation matrix of homogeneous transfer matrix and Newton iteration to obtain the solution of forward kinematics of the space attitude simulation bogie. First, according to the orthogonal relationship between the elements of the rotation matrix and the length constraints of the hydraulic cylinder rods, the mathematical model with nonlinear equations was established. In this model, the elements of the homogeneous transformation matrix were taken as the unknowns, and the number of equations was equal to the number of the unknowns. Then, the approximate solution was obtained with Newton-Raphson method. Finally, the corresponding position and pose were solved on the basis of the relationship between the elements of the homogeneous transformation matrix and the position variables. The method avoids the iterative calculation of trigonometric function with Newton-Raphson method, improving the speed to solve the forward kinematics. Also, the method of forward kinematics of the bogie 6-DOF platform provides the possibility of realizing the space attitude simulation of bogie. It provides the theoretical basis and technical support for the pose solution of other forms of 6-DOF platform.

Based on the existing reliability analysis of bridge structure system, the Markov process steady-state probabilities are used as the reliability assessment indexes of bridge system. The load-sharing process in bridge working condition is analyzed by modification of the system reliability relation matrix. Laplace transforms of the bridge system reliability function and state equations are conducted. Then state equations obtained after the modification of the system reliability relation matrix are solved to deduce the calculation method of Bridge System Mean Time To failure (BSMTTF). A case study is carried out, which indicates that the BSMTTF is closely related to the component target reliability indexes. The higher the bridge system reliability index the longer the BSMTTF, and the higher the component target reliability index the more obvious the increase in the BSMTTF.

A new method, called segment equilibrium method, is proposed based on yield line theory to calculate the limit load-carrying capacity of one-way concrete slabs at large displacement. In this method, the vertical component of steel forces at the hinge-line section, which is considered as the explanation of tensile membrane effect in the one-way slabs, is incorporated into the equilibrium equation of each rigid body divided by the hinge-line. To validate the method, six one-way slab static loading tests with three types of boundary conditions were conducted. The load-deflection curves, ultimate bearing capacity and failure patterns were obtained. The calculation results by the proposed method are in good agreement with the experimental results. So this method is applicable for engineering design.

Rough terrains were prepared using quartz sand with six average particle sizes in a mini-size bin. A special instrument equipped with distance laser sensor was used to collect the digital information of the surface morphology of the rough terrains, and analyze the pretreated data. The fitting curves of the average Root Mean-Square (RMS) height and the average terrain energy as functions of the mean particle size, the physical meaning of the areas enclosed by the frequency axis and the auto power spectrum density curves indicate that the average terrain energy and the terrain roughness have obvious relationship. Thus, the terrain energy can be used to analyze the terrain roughness quantitatively. The analysis of the linear fitting result of the double logarithmic auto power spectrum density curve shows that the slope of the linear fitting line increases with the terrain roughness. It proves the reliability of using the relative positions of linear fitting lines and the slopes to evaluate the roughness of the terrain.

This paper describes the principle of braking energy recovery and release of Hydraulic Hybrid Vehicle (HHV) and analyzes the braking process. The HHV AMESim model is established based on the vehicle dynamic model. According to the actual conditions, the vehicle related hydraulic parameters are set to simulate the process of braking energy recovery and release. Experiments of typical operation conditions were carried out on an experimental bed. Simulation results indicate that the energy recovery rate is 38.7%, and energy release rate is 85.1% after the vehicle overcoming the rolling friction, mechanical friction and the heat loss of the accumulator charge and discharge. The experimental results are in good agreement with the simulation results, indicating that the proposed model can be used to analyze the dynamic performance and the braking effect of the HHVs intuitively. In addition, it provides a reference for future investigation and optimization of HHVs.

Computational Fluid Dynamics (CFD) method is employed to compute the temperature field of open-type hydrodynamic retarder with an effective diameter of 468 mm. By simulation and analysis, the relationship between temperature and rotating speed of the hydrodynamic retarder in the condition of full filled liquid is obtained. The internal temperature field of the hydrodynamic retarder is analyzed. On that basis, the open-type hydrodynamic retarder under the conditions of the same rotating speed and different liquid filling rates are computed. The relationship between temperature and liquid filling rate is obtained. Through simulating, it provides an important theoretical basis for the heat exchange of open-type hydrodynamic retarder at different gears.

On the base of analyzing the working principle of wind turbine with hydrodynamic speed adjusting system, the mathematical model of the transmission system of wind turbine was established. Then the parameters of the transmission system were optimized. The corresponding control system was designed and simulated. Simulation results demonstrate that wind turbine with hydrodynamic speed adjusting system can respond to the change of wind speed rapidly and ensure that the wind turbine runs efficiently, reliably and stably.

The rheological and phase transitional dual characteristics of magnetorheological fluid were studied and its rheological principle was also described. Using ANSYS finite element method, simulation analysis of the torque transmission model was conducted. Choosing magnetorheological fluid SG-MRF2035, a magnetorheological torque servo device prototype (gap of 1.5 mm) was developed. A performance testing platform was built to characterize the torque output of the prototype. The testing results show that the device can provide flexible output torque and the torque is continuously adjustable. This study provides a new way for the wide application of magnetorheological fluid.

The gear dynamic equation was constructed, which takes the tooth surface friction into consideration. The equation was solved by the multi-scale method. The influence of the friction coefficient on the main resonance frequency response was analyzed. On the basis, the influences of the static load, dynamic load and the damping coefficient on the main resonance frequency response were investigated. The results show that the tooth surface friction can effectively reduce the resonance amplitude. When the static load is larger its influence on the resonance frequency is greater. The dynamic load and damping factor can reduce the resonance amplitude, but their influence on the resonance frequency is not obvious.

The vibration response equation of an infinite plate with quadrate steel beams under point force load is derived using mechanical mobility method. The effects of different layouts of quadrate steel beams on plate vibration are investigated by numerical calculation. The results show that the vibration level difference of the target area decrease with the increase in the distance from the load to the target area;the vibration of the target area that is far from the load and near the symmetry axis is amplified for plate with single quadrate beam or several parallel quadrate beams. The symmetrical characteristic of the vibration distribution in infinite plate is remodeled by several non-parallel quadrate beams. The vibration suppression capacity is enhanced by increasing the angle between quadrate beams.

For given ordered data points, the Non-Uniform Rational B-spline (NURBS) interpolation technology was employed to construct the target surface and its intersection curve equations. Based on the interaction of springback effect and double curvature bending influence of sheet metal, the transverse (perpendicular to rolling direction) and longitudinal (along rolling direction) shape control models were developed. The relative positions of control units on the roll's axis and the downward displacement of upper flexible roll were obtained in continuous flexible rolling process. The combination of reverse engineering and continuous flexible forming technology was accomplished. The undetermined coefficient in the transverse shape control model was calculated by means of numerical simulation. Combining with the longitudinal shape control model, verification experiments were conducted. Results show that the final curvature was in good agreement with the target curvature.

An algorithm of 6-DOF robot marking and trajectory planning is put forward, which is used for Multi-point Stretch Forming process of curved surface of sheet metal. First, the drawing area of the curve surface is determined according to the dimensional parameters of the stretch forming machine. Second, mapping software is used to analyze the data points within the parameter domain of the curved surface. Third, the boundary points are extracted through the coordinate relationships of the data points to determine the line drawing domain of the curved surface. Finally, the line drawing domain is projected onto a plane. Simulation results of line drawing operation are displayed by the use of manipulator inverse kinematic theory. Experiment proves that the algorithm can solve the problem of area selection of sheet metal, and provide an effective line drawing method for sheet metal forming.

Taking the phase expansion phenomenon of water freezing as the theoretical basis, this paper experimentally studied the ice adhesion strength on Poly(methyl methacrylate) (PMMA) surfaces with various dent morphologies. Experimental optimization design method was used to arrange the experiments. The results show that, although the dent morphology increases the contacting area between PMMA substrate and freezing water droplet, PMMA surface with certain dent morphology can reduce the ice adhesion strength. Microscopic observation and analysis of the freezing process suggest that the dent morphology on the PMMA surface influences the phase transition on ice/solid interface, which can lead to the reduction of the ice adhesion strength.

In order to design the warm rolling process of medium carbon steel, flow stress under warm compression deformation was investigated by experiments. The experiments were carried out on a Gleeble-3500 thermal simulation machine at temperature of 550~700 ℃ with constant strain rates 0.001~1 s^-1. The influence factors in warm deformation were studied. Results show that the stress-strain curve of medium carbon steel under warm deformation is composed of four stages: the ferrite dynamic recovery, ferrite dynamic recrystallization, cementite spheroidization, and steady deformation state. The main soften mechanisms are ferrite dynamic recrystallization and pearlite dynamic spheroidization.

A reasoning method for ternary projective relations of points on a sphere was proposed based on the analysis of the model and unique properties of the spherical surface. The reasoning method can get the converses and rotations of eight spherical projective relations, and the compositions of the eight spherical projective relations. These results are shown in tables. By querying the tables, the unknown spherical projective relations can be obtained from those already known. This reasoning method completes the spherical projective relational models and lays the foundation for its practical application.

A novel method based on fuzzy theory and Bidirectional Associative Memory (BAM) neural networks is proposed to evaluate the transformer aging problem. In this method, first the features of the transformer are fuzzified to deal with their uncertainty problem, and then fuzzy reasoning and defuzzification are performed to evaluate the transformer aging. During the fuzzifying process, because that the electrical characteristics of insulating oil gas and insulating oil are highly personal relied, BAM neural network is used to realize the fuzzification of these two features. To examine the effectiveness of the proposed method, simulation is carried out with real data. Numerical results show that the evaluations of the proposed method highly coincide with the engineering specifications.

In the process of service flow scheduling based on IEEE802.16d agreement, a good scheduling algorithm should ensure that higher priority tasks get priority service;meanwhile, the occupation of system resources during the scheduling process should be minimized. The advantages and disadvantages of two kinds of existing Earliest Deadline First (EDF) algorithms, i.e. the non-preemptive EDF algorithm and preemptive EDF algorithm, are analyzed in depth. Then an improved semi-preemptive EDF algorithm based on the importance of factors is proposed. In this algorithm the time characteristics, the importance characteristics and the order of reference are taken as the adjustable parameters, and the transmission distance is also taken into account. By simulation, the improved EDF algorithm is applied to the Real-Time Polling Service (RTPS) service flow scheduling of IEEE802.16d agreement. The experimental results show that the improved EDF algorithm has a better balance between advantages and disadvantages than non-preemptive and preemptive EDF algorithms, with smaller and more stable delay.

In this study, first, by executing the distortion correction and calibration on the camera and projector, the corresponding homography transformation matrix is determined. Then, in order to quickly obtain high precision code values from color encoded structured light under complex projection conditions, a color structured light coding method based on topology structure and local structured model is proposed. Finally, the methods for extracting encoded feature points based on waveform analysis and for classifying colors based on clustering analysis are developed. In addition, the characteristics of the encoded images are also exploited to improve the decoding accuracy. According to the homography transform matrix and the matching relationship of the code values of the feature points in the encoded images before and after modulation, the three-dimensional information of the feature points is calculated. Experimental results show that the decoding accuracy is higher than 99.5%. Because that the positions of the projector and camera are not strictly constrained, the method is applicable for building a reconstruction system that can rapidly and accurately generate point clouds from three-dimensional objects.

Based on the idea of virtual point lights used by instant radiosity, a novel algorithm for approximate calculation of the indirect illumination of dynamic scenes was proposed. First, the view-sensitive sampling strategy of virtual point lights and the interleaved-sampling based method of indirect illumination gathering were designed, and their Graphic Processing Unit (GPU) implementation was further presented. Second, the method for reusing and updating virtual point lights among consecutive frames was suggested. Based on this method, the validity testing, reconstruction and paraboloid shadow-map update of virtual point lights were realized. The proposed algorithm treats the dynamic scene as a whole and does not need to obtain the information with respect to static and dynamic objects in advance. Experiment results show that the algorithm can significantly reduce the average computation time of indirect illumination of dynamic scenes.

To cope with the insufficiency of single-attribute features in anti-aggression and in detecting software theft, an algorithm to detect software piracy is proposed based on classification of multi-attribute features. In order to detect software piracy with better accuracy and lower overhead, after semantics-preserving transformations of the software, the algorithm extracts features by the following steps: disassembly analysis, extraction and selection, segmentation and metrics, construction of classifier according to similarities. The algorithm is described and its implementation is depicted. Experiment results show that the algorithm is feasible in improving the robustness and credibility of identification, and resistance to attacks. It overperforms existing similar methods.

First, by introducing a driven Jacobi matrix, the dynamic equations of the hydraulic rigid-flexible manipulator were derived using Lagrange principle and the assumed mode method. Then, a joint angel compensation algorithm was proposed by considering the characteristics of flexible arm. In order to weaken the effect of parameter perturbation and external interference, an adaptive sliding model controller was developed to simultaneously control the rigid motion and suppress vibration with uncertainties estimated by the adaptive law. Simulation results were presented to demonstrate the effectiveness of the proposed approach.

Under the multi-objective control framework, this paper presents a hybrid algorithm to solve the problem of robust static output feedback control for continuous polytopic uncertain system. To obtain the static output feedback gain, the new hybrid algorithm combines the Particle Swarm Optimization (PSO) and Differential Evolution (DE), and the Linear Matrix Inequality (LMI) method. The proposed algorithm can be applied to solve the optimization problem with Bilinear Matrix Inequality (BMI) constraints. The PSO-DE hybrid algorithm is used to obtain a population of controllers, and LMI approach is used to optimize the performance criterion of the system. Taking a hybrid H₂/H_∞ control problem as example, the detailed algorithm is presented to solve the problem of robust static output feedback control. Simulation results show that the proposed hybrid algorithm improves the convergence rate and accuracy compared with traditional iterative methods and the original DE-LMI algorithm.

A fault-tolerant target tracking algorithm is proposed for multi-sensor system with random sensor faults. The sensor measurement fault is modeled by multi-model approach. The multi-model information filtering is established via embedding the unscented transformation into the information filtering architecture. The probability of the sensor fault model is estimated online by multi-model estimation. The information contribution of each sensor is calculated by combination of probability weights of the sensor fault model. The target tracking accuracy and the identification probability of sensor fault are improved via multi-sensor data fusion. The simulation results show that the proposed algorithm can improve the efficiency of target tracking.

To overcome the climbing weakness of traditional spherical robots, a new spherical robot with climbing link mechanism is designed. This new spherical robot can not only walk in the traditional way as an inverted pendulum, but also climb a steep slope with the new climbing link mechanism. The climbing mechanics model of the new spherical robot is created. Then the climbing processes of both traditional spherical robot and the new robot are simulated using simulation software. Simulation result verifies the climbing mechanics model. The parameters which can influence the climbing ability of spherical robot are analyzed and optimized. A prototype of this new spherical robot named BYQ-X is constructed based on the analysis results. The mechanical structure and motion control system are introduced in details. Finally, the accuracy of the mechanics model and the ability of the climb link mechanism are verified by the tests of the prototype.

The network survivability modeling and quantification methods were studied for comprehensive and efficient quantification and analysis of the network survivability. The Heegaard's failure and recovery model and its state transformation were introduced. Then an improved network failure and recovery model based on Continuous Time Markov Chain (CTMC) was proposed. The improved model includes the failure state model and reduces the number of failure and recovery states. Combined with network performance models, an improved network survivability model based on CTMC was proposed. The survivable state transformation process after failure was simulated. Taking the packet loss state as the quantification factor, a network survivability quantification method based on improved CTMC network survivability model was given, that simplified the calculation of the survivability. Simulation results show that the proposed modeling and quantification methods can quantify the network survivability more comprehensively and accurately.

An optimal combing weight estimation algorithm is proposed. It is applicable for signal synthesis under the condition of correlated noises. Different from the traditional algorithm, the operation of the new algorithm only relies on the auto-and cross-correlation functions of multiple signals, and the solution of a trivial quadratic equation with one unknown. Without computing the middle variables, such as noise variance, signal amplitude and noise cross-correlation coefficient, the computation of the algorithm is very simple. Simulations on QPSK, 8PSK and 64QAM signals, with correlated and uncorrelated noises, show that the algorithm can estimate the combing weights accurately and robustly.

The accuracy of the current interest points detection algorithms are usually influenced by the unstable interest points in the non-interest regions. To overcome this disadvantage, an image retrieval algorithm combining interest point detection and region division is proposed. First, the interest points detection algorithm, combining the feature of Scale-Invariant Feature Transform (SIFT) and Harris (IPDSH), detects the stable interest points. Second, using the spatial positions of the stable interest points, the algorithm divides the image into annular region and convex hull, and calculates the color histogram of pixels in the convex hull and the pseudo-Zernike in the local field of stable interest points in the annular region. Finally, the algorithm retrieves the image with the weighted feature vector. Experimental results show that the proposed method possesses high retrieval speed, simple realization, robustness in rotation and translation, and can reduce the impact of unstable interest points. The image retrieval precision can by improve by 7.0%-15.1%.

The surface electromyographic signal (sEMG) is often complicated, and it is expensive and time-consuming to obtain labeled samples of sEMG, especially when it has to be done manually by experts. To overcome such problems, a new semi-supervised boosting algorithm is used to classify multi-class problem of sEMG in this paper. This algorithm can use a large number of unlabeled samples together with a small number of labeled samples to build a better learner. Most semi-supervised algorithms have been designed for binary classification. The shortcoming is that when extended to multi-class classification these algorithms are unable to exploit the fact that each sample is only assigned to one class. The advantage of the new semi-supervised boosting algorithm used in this paper is that it exploits both classification confidence and similarities among samples when deciding the pseudo-labels for unlabeled samples. Empirical study with six movements of sEMG show that the new algorithm performs better than the state-of-the-art boosting algorithms for semi-supervised learning. It gives large reduction in the number of human labeled samples, high classification results, which has practical significance in sEMG pattern recognition

The exist several problems in current measurement of hepatic functional reserves, such as multiple blood samplings for spectrophotometry, lower precision and longer time taken by manual operation, and possible infection risk. In order to solve these problems, a noninvasive and continuous method based on pulse sepctrophotometry for detection of hepatic functional reserves is presented by setting up and analyzing Indocyanine Green (ICG) dye dilution and clearance curves. While ICG is injected intravenously, two light sources at 940 nm and 805 nm wavelengths are used to implement data acquisition by Photoplethysmography (PPG). The ICG disappearance rate in blood plasma (K value) and 15 min retention ratio (R₁₅ value) as hepatic functional reserves parameters can be determined by calculating time-varying ICG concentration values. Compared with blood sampling method, the mean relative error of this new method is below 5%. The results confirm its usefulness in evaluating hepatic functional reserves for clinical liver removal surgery.

After briefing the development of audio system, this article explains the necessity of sound source extraction. New methods of extraction of sound sources from the original recordings of aircraft cockpit using short-time Fourier analysis/synthesis technology and linear prediction analysis/synthesis technology are proposed. The main theory and extracting steps of the two methods are discussed in detail. Two case studies are conducted to describe sound extraction of extending landing gear using short-time Fourier analysis/synthesis, and aerodynamic hiss extraction in flight using linear prediction analysis/synthesis from original recordings. The case studies and technology development reveal that short-time Fourier analysis/synthesis is more common and applicable to most sound sources, and linear prediction analysis/synthesis is more suitable to noise extraction. The two technologies can also be used for audio system developments of other vehicle simulators. Therefore, they possess important value in engineering application.