A simulation model was built for the selfdeveloped parallel hybrid electric vehicle(HEV) with dual clutch configuration using the software of powertrain system analysis toolkit(PSAT) with the forwardlooking simulation approach. The model was used to perform the optimization simulation of the control strategy for the engine startstop in HEV. The results showed that the optimal fuel economy can be realized when the engine start power is 10 kW/1200 r·min-1 and the engine stop power is 4 kW/800 r·min-1. Under these power settings, the engine and the integrated startergenerator(ISG) motor can collaborate well with each other, the operating mode switching of the powertrain is smooth. The vehicle tests validated the practicality of the developed control strategy.

A multibody dynamic model was constructed for the rider-motorcyele system by means of the software ADAMS to study the handling stability of the motorcycle. A rider steering and attitude controller was designed using the software MATLAB. Based on the ADAMSMATLAB cosimulation, the singlelane change and the steadystate circular driving of the rider-motorcycle system was analyzed. The simulation results indicated that the designed rider steering and attitude controller represents the rider behaviors well and the motorcycle follows the target track precisely. The cosimulation of the rider-motorcycle system can be used to study the handling stability of the motorcycle.

A PSOBP network algorithm was suggested to optimize the automotive suspension based on the study of suspension model and neural metwork. The algorithm uses the PSO learning mechanism to improve the weak points of the BP learning algorithm such as the slow convergence rate and the poor global covergence. A mathematical model was established and the stiffness and damping of the automotive suspension were optimized with the suggested algorithm. The results showed that the algorithm can significantly improve the performance of the automotive suspension and the vehicle ride comfort. The suggested algorithm and obtained test data provied a reference for the further study of suspension system.

The impacts of bus running and stopping on the traffic flow were analyzed considering the real operation state of buses at the road segment. On the case of light traffic, treating the bus as a moving service station, a delay model by buses was established based on the queuing theory and the theory of gap acceptance. On the case of heavy traffic, a delay model by buses was established based on the arrival and service representation of vehicles. A solution example was demonstrated. The obtained results provide a theoretical support for determination of the reasonable travel time functions and evaluation of service level at the road segment.

To deal with the problem of traffic accidents such as vehicle collision during the conscious lane change of the vehicle on the highway under dangerous conditions, a now safety lane change model was established on the basis of the previously studied and developed lane departure warning system. The model considered the typical lane change scenario and more actual behavior that most driver often accelerate their vehicles during the lane change process. A simulation program was developed to analyze the model performance based on the software Matlab 7.0. The established model was compared with the developed by California University and the result showed this model is more practical. It can be used to establish the vehicle assistant driving system for lane change.

A model framework to capture the commuter day-to-day dynamic route switching behavior was proposed based on the bounded rationality theory whose application is mainly embodied in the indifference threshold. To express the experience accumulation process in commuter day-to-day travel and the serial correlation and the contemporaneous correlation across the route selection, an indifference threshold model for the pretrip and enroute switching behavior was built in the form of the multionmial probit model. The driver day-to-day dynamic route switching behavior under different ATIS information was simulated using a driver training simulator and the obtained data were used to calibrate the parameters of the indifference threshold model. The parameter estimation results gave insights into the effects of the ATIS information, the driver individual attributes, the imformation quality and the schedule delay on the commuter daytoday dynamic route switching behavior and provided a useful reference for the further analysis of the driver behavior.

Based on the idea represented by authors to reform the ground temperature distribution configuration through controlling and allotting the load, this work focus on the experimental investigation of heat transfer in the UTES and emphasizes to research the effect, temperature potential, ground temperature variety and distribution. Four typical operational modes on the load of heat resources (boreholes of underground heat exchanger), such as equivalent mode, concentrative mode, intermittent mode and heat shield mode were studied experimentally. Thus, a new method of the higher conservation efficiency of UTES has been pursued by means of controlling every load of boreholes to enhance the practical maneuverability in UTES. The reformation of temperature distribution has been ideated and presented, and the behavior of heat shield has been found to realize a higher efficiency and establish an optimization way for UTES.

The numerical simulation method suitable for the liquefied petroleum gas(LPG) boiler was found by review and analysis of various kinds of the mathematical models of the combustion process in the boiler. Based on the method the numerical simulation was performed for the combustion process of the oxygenenriched LPG boiler using the computational fluid dynamics(CFD) software. The temperature fields, the combustion species concentration fields, the formation levels and the districbutions of NOx under different oxygen concentrations were obtained. The relationships among the temperalure field, the species concentration field and the NOx concentration as well as the combustion efficiency and the NOx emissions versus the oxygen concentration were found. Detailed experimental studies were also performed for the combustion process of the oxygenenriched LPG boiler. The comparison between the calculated results and the measured data showed their agreement by and large. The study is of theoretical and practical importance for the engineering design of the oxygenenriched combustion equipment.

The Microbiologically Influenced Corrosion (MIC) of magnesium alloy was investigated using a special designed corrosion method, which was called solid culture medium adherence corrosion method. Scanning electron microscopy (SEM) and polarization curves were employed to analyze the microbiologically influenced corrosion behavior of magnesium alloy. The results indicated that the appearance of SRB resulted in a rapid pitting corrosion of AZ91 electrodes. The main characteristic of the MIC of the magnesium alloy was pitting corrosion. The metabolism of SRB which involved hydrogenase was the main reason for the MIC of AZ91 alloy. The SRB accelerated the cathodic reaction through the cathodic depolarization.

Hypereutectic Al23%Si alloy was ultrasonic treated using a novel ultrasonic treatment equipment. The effects of ultrasonic power on the morphology and distribution of the microstructure as well as properties of the studied alloy were investigated. The course of crystallization and mechanism of microstrctural refinement for the ultrasonic treated alloy were also discussed. The results show that the ultrasonic treatment can greatly reduce the size of primary Si particulates, make them distribute much uniformly, and it can change the morphology of αAl phase from dendrite to equiaxial grain meanwhile the eutectic phase become courser than it was before. The distribution of hardness was even and the wear resistance was increased by ultrasonic treatment on Al23%Si alloy.

By means of the mixture regression design process, the regression equations relating the joint shearing strength and the wettability of the filler metal to its element compositions were established. The influences of the elements on the joint shearing strength and the wettability were analyzed. The results prove that adding more Ti improves the wettability, but at the same time the joint shearing strength decline; adding appropriate amount of Sn can improve the joint shearing strength and the wettability. After considering all the effects of filler metal ingredients to the melting point, the pasty range and so on, the contribution of Cu-Ni-Sn-Ti active filler metal brazed c-BN superhard materials has been optimized. The result is 10～12wt% for Ti and 3～5.5wt% for Sn.

A series of numerical simulations on the typical multipoint stretchforming(MPSF) processes were carried out, the surface quality and shape error of the MPSF parts were analyzed. The results showed that the smaller the size of the punch is, the better the surface quality will be. The shape error distributions on the MPSF parts with different materials are similar. The larger the curvature of the target shape is, the larger the shape error will be. With the decrease of the punch size, the shape error of the MPSF part decreases. When the size of the punch is relatively large, the shape error decreases sharply, but after the punch element size becomes smaller than a certain value, the shape error decreases more and more slowly.

Precision ultrahigh strength steel EN 10 292 TL4225 and microalloy steel H340 LAD are selected for the study. The intermediate frequency (IF) spot welding process is used. The perfect welded joints by using the appropriate welding parameters are produced. The tension strength and the hardness distribution of the welded joint tested, and the microstructure of the fracture surface is analyzed by scanning electron microscope. The test results show that the welded joints have good microstructure and mechanical performance, no internal defects such as crack and shrinkage hole are found, and the mechanical performance completely meets the actual need of product.

A highprerformance preamplifier and data acquisition system was designed to measure the ultralow and fastchanging current in the quadrupole mass spectrometer. The system uses a composite transimpedance preamplifier, a 4thorder active lowpass analog filter and a successive approximation analogtodigital converter. The realtime control and highspeed largecapacity FIFO data buffer of the data acquisition are implemented with the fieldprogrammable gate assay logic. By analyzing the impact of the gain and the bandwidth on the signaltonoise ratio(SNR) and the full width at half maximum(FWHM), the suitable gain and bandwidth were selected. Experimental results indicat that in the highrate scanning, 30 kHz bandwidth improves SNR and doesn't effect FWHM and signal intenssity. With the increase of the gain, the SNR increases. As the gain increases from 106 to 108 V/A, the FWHM increases from 0.52 to 0.6. When the gain increases to 109 V/A, the FWHM increases to 2 because of low bandwidth of the preamplifier. In lowrate scanning, as the bandwidlth need is low, 109 V/A gain enhances the SNR while mainfaining the FWHM of 0.2.

Based on the theoretical analysis of the industrial frequency induction heating technique, a detacher used for the bearing inner race was developed. A basic model was built for the induction heating detacher by the ANSYS finite element analysis(FEA) software. The electromagnetic field in the detacher and the temperature field in the workpiece during the induction heating process were simulated by computer and the distributions of the magnetic density and the temperature as well as the temperature field history were obtained. The ANSYS simulation proved the existence of the eddy skin effect. The design parameters of the detacher for the bearing inner race were achieved by the coupling FEA analysis of the magnetic field and the temperature field in the detacher. The results of measurement of the induction heating temperatures in the bearing inner race validated the correctness of the computer simulation.

At the phase of original design for hydraulic turbine runner structure，the key question is how to design structural components and consider their fatigue life. According to probability density function of fatigue strength under given life from P-S-N curve and probability density function of stress amplitude for hydraulic turbine runner structure from experimental data, the method of calculating working pressure head under predetermined life and the reliability design method of hydraulic turbine runner structure under preset life are accomplished by stressstrength interference model.

To improve the drying quality of grain and the intelligent level of drying process, an automatic detection and control system for the grain dryer based on virtual instrument and the intelligent field bus was developed. The system can perform the calibration and correction of the moisture sensor, automatic detection and control of moisture and temperature, fault detection and alarm, and predictive control of grain discharge. It is characterized by simple structure, high detection and control precision, strong ability of antiinterference, high automation level, and easy operation. Its practical application showed that the system can satisfy the requirement of grain drying, effective by control the discharge grain moisture content, improve the grain drying quality, and reduce the energy consumption of drying process.

A kinematics model and a dynamics model were built for the geomembrane roll movement in the lining paving machine working on the geomembraneseepageproof irrigation canal. The difference between the ideal angular acceleration of the geomembrane roll and its real angular acceleration under the action of the practical friction force was analyzed. By the theoretical analysis it was found that the geomembrane stretch in the paving process can be avoid through the active synchrouized release of the geomembrane. The paving quality can be improved by the reasonable choice of the driving ratio. The results provide a theoretical reference for the mechanical design of the issigation canal lining paving machine.

Pressures and temperatures in the hydraulic system of the 50 type wheel loader were measured under different operating conditions to solve the overheating problem of the loader working in summer. Through the analysis of the experimental data it was found that the system overheating results from the wrong location of the hydraulic oil cooler, and the hydraulie oil flow rate through the cooler is insufficient to provide a dissipating power of the cooler required by the hydraulic system. Tests of the loader after the rearrangement of the oil cooler in the return circuit showed that the dissipating power of the relocated cooler conforms to the design target, and the total system operates normally.

Aiming at solving the existed problems of the power chuck, the application of hydraulic slipring in the power chuck was introduced. Based on the simplified modal of the oil film face seal in the hydrostatic slipring, the relation of friction torque with variable viscosity was deduced. The curves of the friction torque versus sealgap of the hydraulic slipring under constant and variable viscosities were compared. Experiment was performed to validate the correction of the relation. The obtained results provide some references for solving the oil allocating and cooling problems of the rotating hydraulic cylinder.

A control acheme combining the dynamics disturbance force forward feed with μ synthesis was suggested to solve the strong disturbance force from the dynamics coupliing existing in the hydranlic stewart platform and to eliminate the influence of the hydraulic parameter uncertainties on the locus tracking precision. A frequencydomain control model of the driving leg and a dynamics disturbance force model were built based on the inverse dynamics model of the Steward platform jointspace and the dynamics disturbance force was compensated by a forward feed controller according to the strueture invariance principle. A μ synthesis robust controller was used in the closedloop control to improve the robustness of the system and to suppess the unmodelled disturbance force. The experimental results validated the effectiveness of the suggested control scheme.

A new method, SIMPLEXT, for heat transfer and flow field was proposed. In the SIMPLEXT method, the pressure field was determined by the SIMPLER algorithm, pressure field and velocity field which meet continuity equation and momentum equation can be obtained by the PISO algorithm. The influence of velocity correction on neighbors for velocity correction and the synchrony between source term and velocity field were considered. The reliability and validity of the presented procedure were verified by the numerical simulation of the liddriven square cavity flow, espedally when the mesh becomes 128×128. The convergence of SIMPLEXT algorithm is better than that of SIMPLE.

According to the fact that the moment invariant extracted from the flow pattern image contour can identify the image characteristies with translation, rotation and scale variation, a flow pattern identification method was proposed based on the image moment invariant and the particle swarm optimized neural network. The gasliquid twophase flow pattern images in a horizontal pipe were captured by a highspeed digital video system. The image moment invariant eigenvectors were extracted using the image processing technique of the image morphological contour. The obtained eigenvectors as the flow pattern samples were used to train the particle swarm optimized neural network and thus realized the intelligent identification of the flow pattern. the experiment results showed that the successfully trained particl swarm optimized neural network can identify quickly and accurately 7 typical flow patterns of the gaswater twophase flow in a horizontal pipe with the total identification accuracy of 97.5%, providing a new effective method for noline flow pattern identification.

A 2-DoF dynamics model was built for the rainwindinduced vibrations of the cable strueture with the moving upper rivulet. The nonlinear terms of the original system was expanded into Taylor series up to 5 orders, and the simplified ordinavy differential equations were derived. The system of the rainwindinduced vibrations of the cable structure was analyzed by the multiscale method, and the conditions where the steady solution of the system exists were acquired. the phase diagrams and their histories of the cable and the rivulet were obtained by the numerical method. The computation of the Lyapunov exponent of the system with different parameters showed that there are doubleperiodic, quaxiperiodic and chaotic motion behaviors in this system. The effects of the system parameters on the periodic and chaotic motions of the system were analyzed by using the computer algebra language Mathematica.

The dynamic property of selfanchoring cablestayed&suspension cooperation bridge is studied concerning different restraints including the girder, the assisting pier and the main tower. It is stated that this bridge possesses a rather high free vibration period with closelyrelated vibration modes. The participation of cables, tower and girders in the vibration has increased the torsion frequency which brings about a better resistance against the wind. The size of stiffening beams which are adversely affected by the joint action of horizontal forces from the cables that could lower the flexural rigidity, should be increased. The frequency of each vibration mode increases with the growth of rigidity of the restraints while the primary longitudinal frequency is more noticeably affected by the elastic rigidity of restraint than the other modes. The assisting piers at the end spans will be helpful to improve the rigidity of the structure and increase the frequencies of the vibration modes. The influence of the main tower on the dynamic property is determined by the restraints of foundations to the tower bottom. With ample size of the pile groups, the frequency of free vibration is slightly less than that of the sunk shafts.

With the method of Symplectic elasticity，the flange of T cantilever beam was simplified into plane stress plate. SaintVenant analytic solutions of wide flange were deduced separately under concentrated load at the end and uniform or linear shear load applied to the whole span length. For the first time the closed polynomial analytic results of SaintVenant problem were obtained. Then the shear lag coefficient of a cantilever beam with wide flange was obtained through SaintVenant analytic solutions and FEM. So the method introduced here is proved to be correct，and it makes theoretical analysis of shear lag in T cantilever bridge simpler.

The basis function of the Element Free Galerkin method was researched. It can be shown that the shape functions and their derivatives calculated with orthogonal basis function were equivalent to that calculated with polynomial basis function.The matrix was transformed to a diagonal matrix using orthogonal basis function, which made the computation of an inverse matrix much simpler and less computational time, and had no effect on the calculating accuracy. Two numerical examples were used to prove the validity of the proposed method and comparisons of CPU time were given.

The finite element analysis software, ANSYS was used to perform the numerical simulation. in which some different curve of the interface topography was used. The distribution of the residual stress and the relationship between the residual stress and the interface asperity dimension were obtained.

The implement approach of reanalysis methods for topological modifications of structures was discussed based on IDEAS . The frame of the reanalysis system was designed based on the open solution architecture of IDEAS, and the commands and matrix operation methods were used to write the program of the reanalysis method. The matrices manipulation results stored in database can be reused to save the calculation time of the solution. The presented method provided the convenient way for applications of the reanalysis theory.

A 2 DoF nonlinear vertical vibration model was built according to the real parameters of the 4200 model rolling mill. The numeric solutions of the vibration equations were obtained by a program compiled by means of the Matlab software. The phenomena of the quasiperiodic vibrations and the chaotic movements of the vibration system versus the exciting frequency and the effects of the stiffness variation amplitude of rolled plate on the vibration state of the mill were analyzed by the methods of bifurcation diagram and poincare mapping. The results showed that when the exciting prequency is in the nearby of the eigenfrequency, larger stiffness of the rolled plate causes more obvious nonlinearity of the milling system, leading to system unstability more easily.

The study of the DEADA model with interval numbers and its sensitivity analysis builds the DEADA model with interval number and studies the arithmetic and the regulation using the algorithm interval number and linear programming. Furthermore, we study the sensitivity analysis of the integration model through analyzing the feasible solution of linear programming in order to discuss the disturbance of the upper and lower vertex of interval number. All the studies can be applied to fuzzy pattern recognition such as picture recognition and engineering pattern recognition under the fuzzy condition.

Efficient and abrasionresistant bionics nonsmooth was introduced into the design of the geologic survey bit. The macro nonsmooth ratio of the bionics impregnated diamond bit was optimized by test in accordance with different formation characteristics. The field production tests of the optimized bionics bit were performed in Huadian, Jilin province and their results showed that the bionics bit optimized in accordance with the formation to drill demonstrated higher penetration ability and longer service life than those of conventional impregnated diamond bit.

The coupling of the creatures is a kind of combination or coupling of creatures through multifactors (coupling elements) such as their morphologies, structures, materials and other biological features. It presents multiple functions to the environment and realizes the best adaption for the biotopes. Based on the basicelement theory of the extension theory and the conjugational theory, a systematic analysis is made for the typical organisms by biomimetic engineering, the extensive model of the coupling elements and the way of the coupled response is established in the study. The proposed establishing method of the extensive model by multifactors coupling bionics provides a formalized model for the expression of the systematized information and knowledge of biocoupling. The hierarchical structure model of the contribution evaluation for multifactor coupling elements is also established for typical organisms. The profitability of the coupling elements is quantitatively analyzed by using hierarchy process of the optimal experiment design.

An optical contact angle measuring instrument was used to measure the contact angle of wing surface of 6 species of dragonfly, it was found that the dragonfly wing is hydrophobic with the contact angle 135.14°~149.85°, depending on the species of dragonfly without respect to the measure position. The observation of dragonfly wing by the scanning electronic microscopy showed that its wing surface is consist of nanolevel structure of microdimensional columms, the colurm diameter is 66.90~200.73 nm, the column spacing is 20.00~650.00 nm, the thickness of the wing membrane is 1.58~4.08 μm. The hydrophobic mechanism of the dragonfly wing was analyzed and the results showed it suits with the Cassie model. The wettability of the dragonfly wing is depend on the effect of the nanolevel column structure on the wing surface and the components from the body surface layer. The study provide a theoratical basis for the design of the microaerocraft and development of the selfcleaning material.

The existing IBE system can avoid the problems of PKI, but in practical application it is lacking of identity management, authority management and key management, while it can not resolve the crossdomain interoperability. In this paper, we designed a new scheme: IBE system based on trust service, through add four mechanisms to the existing system to complete the system. The architecture and mechanism of the new scheme is introduced. This scheme validates the superiority of high security, high efficiency and low cost compared with PKI.

With the appearance of Web technique, it becomes possible to develop application programs based on the combination of Web Service, which is the hotspot of research. Based on the principle of Interface Automata (IA), we provide a design method of web service combination in which each web service corresponds to an IA model and use the IA combination process to represent the Web combination. When the requirement is change, the IA will change and IA combination will be built again. Moreover, when the requirement is stable, we can find the way of web combination through searching the path of IA combination. This method is used to rapidly generate new application systems that need the requirement.

A remote sharing solution using the VPN and the SSH technologies was proposed based on X Window system. A X server was built in the remote client to redirect the I/O of the control software. A 3layer network structure was built using the VPN and the SSH technologies to control the visit authority, protect the data transmmision, provide the collaborative services such as file, video and audio. The remote control of the scientific instrument was implemented through remote operation of the control software. The solution was applied to realize the remote sharing system of MAT 900 type mass spectrometer successfully. It can also be used in remote operation of the software of various scientific instruments running on the UNIX/Linux platform.

A twolevel hierarchical scheduling algorithm was proposed to improve the certainty of the hybrid realtime tasks. The structure of the scheduling algorithm was jiven, the periodic realtime tasks are scheduled by adding a scheduling module and the aperiodic realtime tasks are scheduled by the system kernd. A task model was established and the schedulability of the task set and the signal jitter were researched. The proposed scheduling algorithm was applied to the hydraulic machinery control system based on the software Windows CE,NET,and the hardware platform of the control system and the task division were analyzed. The validation of the proposed scheduling algorithm was performed and the results showed that the algorithm can ensure the certainty of the hybrid realtime tasks, and reduce the signal jitter, thus improve the precision of the control system.

A UML design pattern of product configuration knowledge expression was provided. In order to enhance the embeddedness of product configuration, the general part structure is designed into the tree structure, at the same time, the version management is supported. In order to enhance the independent ability of the configuration logic, the design of configuration logic is separated from the general product structure. Not only the configuration logic is easily maintained and the configuration reasoning is predigested, but also some interfaces are provided for the updating and substituting of configuration logic in the future.

A novel software authentication protocol SRMP (Software Right Management Protocol) that dynamically distributes copyright is presented. The protocol effectively implements safe registration and software migrating by separating the kernel from the tryout version, integrating the kernel, the software environment and the version together. It ensures the data security and integrality in the protocol exchange in network through encryption, authentication and digital signature. It realizes the dynamical distributing copyright and solves the hard disk copy attacking by online to the server and register the kernel dynamically.

A computeraided diagnosis system to process the breast tumor ultrasonic images was proposed based on the arisotropic diffusion and the support vector machine. Total 26 features about the textures and the geometry of the tumor images, including the autocorrelation coefficients, the mean square deviation coefficients, and the compact ratios, were extracted. These features not only reflect the essential differences between the malignant breast tumors and the benign ones, but also consider the characters of the medical ultrasonic images as the coherent images. Experimental results showed that the diagnosis accuracy of the system is 93.47%, its sensitivity is 94.3%, its specifity is 92.5%, its positive predictive value is 93.55% and negative predictive value is 93.4%, the area under the ROC curve is 0.9718. The overall performance of the proposed system is superior to that reported in the influential literature\[1\].

A computeraided diagnosis system to process the breast tumor ultrasonic images was proposed based on the arisotropic diffusion and the support vector machine. Total 26 features about the textures and the geometry of the tumor images, including the autocorrelation coefficients, the mean square deviation coefficients, and the compact ratios, were extracted. These features not only reflect the essential differences between the malignant breast tumors and the benign ones, but also consider the characters of the medical ultrasonic images as the coherent images. Experimental results showed that the diagnosis accuracy of the system is 93.47%, its sensitivity is 94.3%, its specifity is 92.5%, its positive predictive value is 93.55% and negative predictive value is 93.4%, the area under the ROC curve is 0.9718. The overall performance of the proposed system is superior to that reported in the influential literature\[1\].

A computeraided diagnosis system to process the breast tumor ultrasonic images was proposed based on the arisotropic diffusion and the support vector machine. Total 26 features about the textures and the geometry of the tumor images, including the autocorrelation coefficients, the mean square deviation coefficients, and the compact ratios, were extracted. These features not only reflect the essential differences between the malignant breast tumors and the benign ones, but also consider the characters of the medical ultrasonic images as the coherent images. Experimental results showed that the diagnosis accuracy of the system is 93.47%, its sensitivity is 94.3%, its specifity is 92.5%, its positive predictive value is 93.55% and negative predictive value is 93.4%, the area under the ROC curve is 0.9718. The overall performance of the proposed system is superior to that reported in the influential literature\[1\].

By reasoning the conditional independence and dependence between attribute values based on General Information theory, redundant attributes are removed naturally and a submodel is constructed for each test sample rather that the whole sample space. The joint probability density and conditional probability is estimated based on marginal computation. The experimental study on the UCI data set shows that, this algorithm can describe the marginal dependency of mixedmode data more intuitively and accurately.

Two classes of mixed integer bilevel programming problems were discussed. For the first class of problems, the follower’s programming which has no integer variable is convex . In the other class of problems, the follower’s integer variables were involved in the follower’s functions, and these functions are polynomial in follower’s integer variables and linear in the follower’s continuous variables. A new genetic algorithm was proposed for solving the mixedinteger bilevel programming problems. In the proposed algorithm, a new crossover operator was designed based on the simplex evolutionary technique, in which the best individuals were used to generate better offspring of crossover. The simulation results illustrate the designed crossover operator is better than that used in traditional genetic algorithm.

A modified method named MSTNR in DCT domain was proposed by uisng MMSE technique to estimate the a priori SNR at current speech frame at the second step in TSNR with the assumption of Gaussian model in the signal. The proposed approach is simple and requires no any a priori condition on the gain function of the speech enhancement system. And it can remove the drawback of the TSNR method while deep its advantages. Experimental results show the improved performance of the appraoch under different noise conditions.

TOA(Time of Arrival) estimation is investigated for ranging applications with UWB signal source. TOA estimation in the frequency domain is proposed, which reduces the sampling rate and the power consumption of receiver, as well simplifies the receiver structure. The computational complexity and performance of both algorithms based on the state space are compared. The simulation analysis through the practical channel model shows that boty of the TOA frequency domain algorithms can achieve precise location estimation, but the computational complexity of normal power method is lower. The proposed methods particularly suit to the positioning applications that need low rate and low power consumption.

Based on a uniform linear array of electromagnetic vector sensors (EMVSULA), a new algorithm was proposed for twodimension (2D) direction of arrival (DOA) and polarization of the wideband cyclostationary signals embedded in colored noise under a multipath environment. By sufficiently utilizing the cyclostationarity of the signals, the proposed method not only resolved the problem of 2D DOA and polarization estimation of wideband sources, but also performed signal selective DOA. The proposed algorithm had the ability to suppressing additive colored noise and wideband interference. Simulations illustrate the efficacy of the algorithm.

A new method based on higherorder spectrum was proposed to analyze the differential AE signal in this study on the nonGaussian characteristics of the differentia1 load of concrete．The nonGaussian characteristics of differential AE signal of different load were extracted by using bispeetral method．The average bispectral magnitude of differential AE signal was taken as an eigenvalue for the quantitative analysis of the deviation degree from normality．Meanwhile，the signal bispectral features and nonGaussian characteristics of different load were discussed．The experimental results show that the extracted eigenvalue reflects the deviation degree from normality and that the transition of load can be determined from the extracted eigenvalue．However, the Highorder spectrum and correlation of the AE have been presented more sensitive than normal parameters, such as AE events and RMS.

High Frequency (HF) channel is timevarying multipath fading channels involving multipath fading, Doppler frequency shift and spectrum spread. Based on the usually used Watterson model, an improved software channel simulation is proposed. The time delay and frequency shift of each propagation path are both modeled as Gaussian random variables, making the improved Watterson model applicable to more HF propagation channels with larger bandwidth. Then the delay power profile of wideband HF model is sampled and added into the channel simulation and a dynamic software simulation of wideband HF channel is implemented using the tapped delay line structure of Watterson model. The new model is implemented by a Matlab program and proved to be valid. The proposed model of HF channel can easy to be implemented in realtime hardware channel simulator.

DSP Builder, a powerful FPGA design tool, is utilized to realize a chaosbased communication system. First, the timing simulation is performed in Simulink. Then, RTL simulation is performed in Modelsim. Finally, resource consumption and timing restriction of the system are analyzed. The result of hardware realization is consistent with that of computer simulation. Also, the hardware system can be easily extended.

The original twostage Kalman filtering algorithm was simulified according to the system characteristic and the adaptive filtering was introduced to avoid the divergence of fillering, thus a new improved algorithm was achieved. The theoretical analysis and the simulation results showed that the calculation volume of the new algorhtm is less than that of the Kalman filtering with extened dimensions, and the new algorithm can avoid effectively the filtering divergence.

It is hard to build a mathematical model exactly for the exoskeleton suit because it is a multirigidbody, multi DoF and nonlinear system. Using the BP neural network to simplify the model building procedure, a dynamic model was built for the swing leg of the exoskeleton suit. A neural network sensitivity amplification control was proposed and its feasibility was validated by simulation based on the Simulink and SimMechanics toolbox of Matlab.

A dynamics model and an energy consumption model were built for the twowheeled selfbalanced robot to use effectively the limited battery capacity and extend the operation time of the robot. Because the battery energy function has the exponential form, it is difficult to derive an analytic solution. So a stepped asymptotic optimization strategy was proposed based on the improved genetic algorithm. The velocity and the energy consumption of the robot were optimized by means of the global optimization technique, solving the minimization problem of the energy consumption of the robot. The effectiveness and the realtimeness of the algorithm were proved by the simulation.

A stereovision servo system was proposed based on the parallel processing technique for realtime tracking and grasping the moving object. In the system, a fast template matching method was used in combination with a fast reliable stereovision matching method to achieve the image matching and identification. A Kalman filter was applied to predict the position and speed of the object. The parallel processing techuique was adopted to enhance the image matching speed. The image identification results were transmitted directly to the robot and the stereovision platform controller to achieve the visual sereo of the robot and the stereovision platform to realize the tracking and grasping functions. The computer simulation and experiments on the system were performed and the results showed that both the tracking precision and matching speed were improved