The zinc nickel alloy depositions with grain size of 14\_33 nm were electroplated by using adequate additives in an alkaline bath. The microstructures and phase compositions of the depositions were analyzed by TEM, SEM and XRD technologies. It was demonstrated that the depositions obtained from the bath with current density range of 0.5\_50 A/dm2 are all single γphase (Ni5Zn21). Crystal size decreases from 33 nm to 17 nm with the increase of current density in low current density range, and tend to stable in the range of 14\_17 nm in medium and high current density range. There are many clusters of grains, whose sizes increase linearly with the increase of current density.

A kind of low alloy steel was designed by adopting the orthogonal method whose main compositions were silicon and manganese elements. In order to improve the hardening ability of low alloy steel, a small amount of chromium and boron elements were added. Moreover, a small amount of rare earth and titanium elements were used to modify the microstructure. Under the condition of the air cooling the structure of this steel is bainite and its tensile strength, hardness, impact and fracture toughness excel 2000 MPa, 55 HRC, 30 J/cm² and 70 MPa·m^1/2 respectively. In the pin abrasion and impact abrasion tests, the air cooling bainitic cast steel presents excellent abrasion resistance. The production process of bainitic cast steel is simple and the cost is low. When it is used to make the jaw plate and hammer of crusher, the servicing life of bainitic cast steel is 1-6 times higher than that of high manganese steel.

The effect of Cu content on the morphology of intermetallic compound Cu₆Sn₅ formed in the Sn-Ag-Cu/Cu interface and solder joint shear strength was investigated using SEM, EDS, TEM and XRD analyses. The results revealed that a scalloplike intermetallic compound Cu₆Sn₅ was formed in the interface of Sn-Ag-Cu solder joint. The morphology of Cu₆Sn₅could be controlled by changing the Cu content and correspondingly the large needle Cu₆Sn₅ can be avoided, giving rise to increasing the shear strength of the solder joint with a dominantly ductile fracture. The results of aging experimentation showed that under the aging condition of room temperature and 1000 h, morphology of Cu₆Sn₅ was transformed into long and hollow hexagonal faceted needle in the Sn-Ag-Cu solder with high Cu content, which made the interface between the Sn-Ag-Cu and Cu into the weak section, decreasing the shear strength.

The following three forming technology modes were simulated by an explicit finite element analysis software: ①the titanium alloy retiary sheet is contacted with the press punch directly, ②the titanium alloy retiary sheet is sandwiched by polyurethane elastic cushions and contacted with the press punch indirectly, ③the titanium alloy retiary sheet is sandwiched by steel paddings and contacted with polyurethane elastic cushions and the press punch indirectly with the blankholder device. Numerical simulation results showed that dimples and wrinkles occured on the surface of the titanium alloy cranial prosthesis in the first two forming technology modes. These two defects were suppressed when polyurethane elastic cushions and steel paddings were used between the titanium alloy retiary sheet and the press punch with the blankholder device in the forming technology mode, and the 3D curved surface of prosthesis with good quality can be acquired. The experimental results showed that the titanium alloy cranial prosthesis can be eligibly shaped with the multipoint forming technology under the appropriate forming process.

The particles of Z type Hexaferrite Ba_3-xLa_xCo₂Fe₂₄O₄₁doped with Lanthanum were synthesized through a sol-gel method. The maximum mole fraction of La₃₊ was determined and electromagnetic property and microwave absorbing property of particles were studied. The maximum mole fraction of x for La₃₊ is 0.30 . The magnetic properties are improved due to the addition of La₃₊ ，and the microwave absorbency is better as the increase of La₃₊ amount from 2.0 GHz to 10.0 GHz for the series of particles doped lanthanum.

In a cell without separating the cathode and anode spaces, the precursor NiTim(OR)_3m+1(acac)_m+1 was synthesized and was directly hydrolyzed to prepare nano-crystalline Ni/TiO₂ powder by a sol-gel process and calcinations at 550 ℃ for 2 h. The high active nanoNi/TiO₂ modified electrode was prepared by electrochemical synthesis and electrodeposition. The cyclic voltammetry were adopted to study the redox behavior of the nanoTiO₂ and nanoNi/TiO₂ electrode in H₂SO₄ solution. Results showed that the precursor contained acac group and nanoNi/TiO₂ of 20 nm were obtained. There were two pairs of welldefined redox peaks for the nanoNi/TiO2 electrode in 1 mol/L H₂SO₄. The discharge current of nanoNi/TiO₂ electrode increased obviously and reached to 75 mA· cm^-2.

The framework of simulation models library was developed to satisfy the requirements of analyzing the model of vehicle dynamic control. The function partition and the objectoriented hierarchy structure of framework were introduced in detail. Based on MATLAB simulation platform, the reusable technology was applied to construct the model library of vehicle components. The S function blocks of all subsystems of vehicle were programmed with language C/C++. The 3D virtual reality of vehicle simulation supported by VRML tool box of MATLAB was realized. The application examples demonstrated that the proposed simulation model library for vehicle components could be referenced by other applications.

In order to reveal the effects of the longitudinal spacing between the vehicles and vehicle number in platoon on the aerodynamic drag coefficient of vehicle platoon, the numerical simulation method was used to study the aerodynamic characteristics of 11 cases of a serial vehicle platoon in the intelligent transport system. Results indicated that the drag of platoon decreases with the decreasing of the vehicle spacing at fixed vehicle number. At fixed longitudinal spacing, with the increasing of vehicle number in platoon the average drag coefficient may decrease by 20%_30% and the vehicle with the lowest drag is about at the center of the platoon.

To meet the requirements of fuel cell vehicle data acquisition and monitoring, an on board data monitoring system was developed. The hardware is constructed with embedded PC, communication card and field bus data acquisition module. The monitoring software was developed by the embedded operating system XPE in the virtual instrument development platform. The fuel cell vehicle CAN bus and sensor channel data acquisition, data storage, information monitoring and malfunction alarm were realized by the developed system. The system provides a powerful tool for research and development of the fuel cell vehicle.
Key words：

Summarizing the existing prediction models and considering the characteristics of the traffic itself such as nonlinearity, complexity and uncertainty, a shortterm traffic flow prediction model based on support vector machine(SVM) was proposed. The real traffic data validation and the comparison with the traffic flow prediction model based on the BP neural network resulted in a conclusion that this model is superior to the BP neural network model on the aspects of prediction accuracy, convergence time, generalization ability, optimization possibility and so on.

Gas chromatography (GC) technology was used to analyze the ethanol and acetaldehyde in engine exhaust gas. Three sampling methods were compared including direct sampling, bag sampling and absorption sampling. An analysis system was established which use a gas chromatograph as analyzer and a twostage absorption advice as sampling system in which the deionized water in the ice water bath was used to absorb the ethanol and acetaldehyde.Ethanol and acetaldehyde are successfully separated and measured by the system and the detection limits are 0.19 μg/mL and 0.18 μg/mL respectively. Separation and measurement of the ethanol and acetaldehyde in actual ethanol/diesel engine exhaust are realized.

The fatigue lifetime distribution pattern of the automotive chain 06BT was studied experimentally. The results showed that its lifetimes under high stress test conditions comply with Weibull distribution with two parameters, the form parameter and the measure parameter which augment with the test load. The sampling test by accelerating test at load level 0.30Q can save time 14% in comparison with that at normal load level 0.25Q, and the test at 0.33Q save time 25% than test at level 0.25Q. The SEM photographs of the break surfaces of the chain showed that the surfaces broken by load level 0.25Q are smoother than those by level 0.40Q.

To improve the safety and robustness of automobile components, the reliability robust optimization theory and multiobjective decision making methods are proposed for automobile frontaxle to change the reliability robust optimization problem into the multiobjective optimization problem. And the global best and personal best of particle swarm algorithm are selected by the grey relevancy theory. Further more, the grey particle swarm algorithm is presented for multiobjective models solution in the reliability robust optimization design. Comparing with traditional methods, the reliability robust optimization design information of automobile frontaxle can be obtained accurately and quickly by the proposed method.

Based on the construction analysis of the mechanical structure of the rotary actuator, a mathematical model was built for the precision rotary actuator with the PZT stack as actuating element. The developed actuator prototype was tested and its angular resolution, rotating stability, rotating speed and carrying capability were measured. The results showed that the actuator prototype works stably and reliably, its rotating speed can reach 2990.6 μrad/s, the torque output is 0.147 N·m, the minimum stepping angle(angular resolution) is less than 0.32 μrad, and the rotor can rotate around the axis for 360° continuously.

A safetycab type rollover protective structure(ROPS) was developed for the wheelloader ZL80G, and the computer simulation and the test methods of the plastic limit behavior analysis of the whole loader class ROPS with the lateral loading were put forward. The simulation and test of the ROPS prototype were performed to investigate the mechanism of the lateral deformation, the loading behavior, and the energy absorption of the ROPS. It appeared that the simulation results were in agreement with those of the test as follows. The plastic hinges emerge at the design positions of both ends of the lower and upper crossbeams successively. The lateral plastic limit loading capacity is 1.27 times of the standard value. The ROPS absorbs 72.1% of the total rollover impact energy during its yielding. The load capacity well matches the energy absorption capability. The developed ROPS of the wheelloader ZL80G meets the requirement of standard ISO3471 to the lateral loading capacity.

Based on the former drag reduction experiments of the bionic nonsmooth body of revolution models in the lowspeed wind tunnel, the simulation wind tunnel tests at the wind speed of 44 m/s were performed for the body of revolution models with the tail rectangularly disposed by three different nonsmooth structures, such as the convex dome form with the diameter/height of 1.0/0.5 mm, the dimple concave form with the diameter/depth of 1.0/0.5 mm, and the riblet form with the width/height of 1.0/0.5 mm, in contrast with the body of revolution model with smooth surface. Test results showed that the bionic nonsmooth structures on the tail of the models can obviously reduce the drag due to the pressure drop, especially for the model riblet nonsmooth form, its pressure drag was reduced by 25.3%, and its total drag was reduced by 11.12%.

To minimize the control error caused by the long time delay in the production process of the asphalt concrete mixer, a method to predict the signals with long time delay by means of the digital signal processing techmique was proposed. A linear filter required by the prediction was designed to perform the prediction of the temperature change history in the production process of the mixer. Combining the prediction control with the fuzzy control, the centralized and intelligent control over the production process of the mixer was realized. Test results showed that the fuzzy prediction control improves greatly the accuracy of the control system and the working reliability and stability of the asphalt concrete mixer.

Aiming at alleviation of the problem associated with the strong forcefeed back impact and the poor position tracking in the bilateral hydraulic servocontrol system with the parallel forcefeedback, an improved bilateral servocontrol algorithm with parallel forcesense was proposed. The algorithm takes the error betwen the master's manipulative force and the slave's resistance as the signal to control the driven end, and the position and force errors form the master's driving movement. Based on the masterslave control features of the teleoperated engineering robot, its dynamics model integrating the proposed algorithm was formed and an experimental system for the 2DOF bilateral servocontrol was built and the experiments were performed and analyzed. The results showed that the proposed algorithm improves the manipulative capability of the masterslave control system, the operator can feel the teleexistence of forcesense, and the system makes a feature of control simplicity, quick convergency and good realtime character.

The robust stabilization of uncertain systems with constant input delay was discussed to obtain an improved reduction method and controller design scheme for the systems based on LMI approach since the existing reduction methods were often complex and difficult to implement for the robust controller design. The designed robust controller by using this proposed method proved to be simple and easy to realize. Simulation results and numerical examples demonstrate the effectiveness and superiority of the method.

The optimal control design for the uncertain piecewise linear system has been converted to the problem of optimizing upper bound and seeking lower bound of the optimal control performance. Therein the optimizing the upper bound of the performance can be expressed as a problem of bilinear matrix inequalities(BMI) in which the feedback gain is taken as the optimization parameters, and the seeking the lower bound is a semidefinite programming problem based on the linear matrix inequalities(LMI). A mixed algorithm combining the chaos optimization and the interiorpoint method was designed to solve the BMI. The effectiveness of the control design and its solving algorithm was demonstrated successfully by a numerical example.

A 2DOF dynamic model for the working device of a hydraulic excavator was built, and a state observer was designed according to the sliding mode control theory on this basis，a variable structure controller(VSC) with the sliding mode observer for the excavator working device was developed by means of the reversal procedure. The tests were performed on the modified excavator to compare the performance of the bucket tip to track the prescribed straight line under the control of the developed VSC with that of the conventional PID controller. The results showed that the errors for PID controller are more than 40 mm, but those for VSC are less than 30 mm. It shows that the proposed VSC is characterized by good dynamic performance for tracking the prescribed line, and robustness against the uncertainties in the system.

In order to eliminate the noncoded design and control links in the development process of the graphic virtual instrument softwares, a program autogeneration mechanism for the virtual instruments was developed. The mechanism bases on the modularization of the needed functions in the professional softwares, builds centralized relationships among modules and module sets. The control links were replaced by the data flow between modules, the design links were replaced by relationship match between modules, and the autogeneration of the professional software on demand was realized. The feasibility of the program autogeneration mechanism was verified by means of a selfdeveloped software development platform.

A target code generation technique based on GDAG and register states was presented. Active information was added to the general DAG, which was denoted as GDAG. Active information together with register states were used during the target code generation. Since information about registers is utilized effectively target codes of high running speed can be generated.

In order to reduce the complexity of the state space in korder Markov predictor, a 2step Markov predictor was proposed to make path prediction over WLAN. Its feasibility was proved by calculation conditional entropy of 2step and korder Markov predictor. In addition, the analysis and comparison of the predictor accuracy for two kinds of Markov predictor using a twoyear trace of the mobility pattern of 1200 users on Dartmouth's campuswide WiFi wireless network were conducted. The results show that the proposed Markov predictor is more stable than oneorder Markov predictor in different length trace files and can also dramatically reduce Markov state space complexity with almost same prediction accuracy of twoorder Markov predictor and higher accuracy than one order Markov predictor.

A new method for the anomaly detection based on the attributes similarity and the cloud model was proposed to alleviate the high false positive rate problem in the detection. The method for calculating the attribute similarity and the attribute weight was given based on the different contribution of the attributes to the classification in combination with the concept of the similarity between the data object space and the attribute space. The method can reduce the dimensionality of the network data space, improve the veracity of the anomaly detection. The experiment results indicated that the proposed method is characterized by that few a prior knowledge is needed and the parameters are easy to determine. It can enhance the detection performance and reduce its false positive rate.

The semiMarkov decision processes(SMDPs) were studied by the Mstep lookahead policy iteration(PI) based on the performance potentials. A Mstep lookahead PI algorithm based on the solution of the performance potential theory was proposed. The algorithm can be used to the standard PI as well as the conventional asynchronous PI, and is also consistent with the SMDP optimization under both discounted and averaged criteria. The formulation for the Mstep lookahead PI based on TD learning under both performance criteria was given. The features of the above algorithm were demonstrated by a numerical example.

A combination method of empirical mode decomposition (EMD) and harmonic wavelet transformation were used for harmonic signals extraction in chaos.Harmonic wavelet transformation to decompose the observed signal into several narrow band signals was conducted first. Then each narrow band signal was decomposed into a number of intrinsic mode function components (IMFs) which was selected according to power spectrum dense, and the harmonic signal was extracted. Performance of the method was evaluated by both a noise reduction factor and correlation between the extracted signal and the original noisefree signal. Theoretical analysis and computer simulation show that this method is effective in extracting harmonic signals provided that signal to noise ratio is not very low.

The traditional training method of Gaussian mixture model (GMM) is sensitive to the initial model parameters and easy to lead to a suboptimal model in practice. To solve this problem, a hybrid algorithm was proposed to optimize GMM model parameters globally and to improve parameter precision by using genetic algorithms (GA) and fuzzy clustering estimation method. An adaptive mixture crossover rate and mutation rate of GA and fuzzy minimum objective function (FMOF) algorithm were used to form a hybrid operation for reestimation of the model parameters, which can improve GA convergence speed. Experiment results based on PKUSRSC database show that this algorithm can obtain more optimized GMM and better than traditional GMM training method and maximum simulation training methods.

An algorithm for jointly estimating range, frequency and DOA of near field narrowband polarized sources with a uniform linear array (ULA) of orthogonal dipolepair was presented. Without spectral peak searching and paring among parameters, the proposed approach can be employed to arbitrary Gaussian noise environment since the fourthorder cumulant is employed. The simulation results show that by exploiting the source polarization, the proposed method outperforms the conventional method where the polarization information is absent.

In order to overcome the disadvantages of low accuracy, high complexity, and lack of robustness in many actual pitch detection algorithms, a new algorithm for pitch detection on the basis of local modulus maxima of dyadic wavelet transform was presented. Simulation results indicate that this algorithm possesses better pitch detection precision for the speech signals under strong background noise, low calculation complexity, high resolution, and capability for real time implementation.

An image denoising method was proposed based on a local statistical model of wavelet coefficients. This method modeled the distribution of wavelet coefficients as an adaptive Gaussian mixture model. This model took into account intrascale dependencies between wavelet coefficients and it was adaptive to the wavelet subbands corresponding to three orientations in the image. Based on this model in a Bayesian framework, a spatially adaptive Bayesian shrinkage function was obtained and each modified coefficient was decided separately. Experimental results demonstrate this method improves the denoising performance and preserves the details of the image.

Properties of channel correlation in frequency domain were used to simplify channel estimation algorithm in MIMOOFDM system. Both FFT orders and times of matrix inversion were decreased greatly. The decreased value can be determined by channel properties. Several channel models were also studied to simulate the algorithm's performance. Simulation results show that, for channels with good correlation, the decreasing in performance is very small.

To alleviate the difficulties in the detection and recognition of the moving objects, even the possibility of the object misclassification, due to the effect of the variation of the moving object shadow and the background factors, a mixed Gaussian background update model based on the object level and a moving object shadow detection algorithm based on the RGB color variation degree were proposed. It performs the mixed Gaussian background update according to the object highlevel semantic expressions, such as movement segmentation, object recognition, Kalman movement tracking, etc., in the light of spatio temporal features of the pixels, eliminates the effect of the prolonged traffic standstill due to the traffic control signs and the rushtime traffic congestion on the background extraction in the mixed Gaussian model based on the pixel level, avoids the object misclassification due to effect of the moving shadows. The results of experiment on the video pictures of different traffic conditions showed the proposed technique is robust and selfadaptive.

After study of the fueloptimal lowthrust transfer with variable thrust and analysis of the difficulty of numerical algorithm for traditional trajectory optimization model, a trajectory optimization model, or nominal orbit optimization model was given by the introduction of nominal orbit concept, which can overcome drawbacks of traditional trajectory optimization model with rapid convergence and easy initial guess. An example of lowthrust EarthMars orbit transfer was presented in order to demonstrate feasibility of using this nominal orbit method and main applications and their scope were given.

In consideration of an efficient approach for autonomous relative orbit determination and in order to reduce communication requirement and calculation burdens onboard for satellite formation flying, a “Microwave+Laser measurement” strategy and decentralized architecture were chosen and decentralized SchmidtKalman filter was used for the relative orbit state estimation. Simulation results verified the validity of this navigation method.

In order to design a low cost high performance digital overload control system for winged missile by common rudder described by first order plant, on the basis of the assumption that the single channel of the missile is a third order plant, the dynamics of the missile were separated into overload subsystem (low speed subsystem), angular rate subsystem (middle speed subsystem), and angular acceleration subsystem (high speed subsystem). And AdamsBashforth numerical approach was used to discrete the subsystems for the design of a discretetime dynamic inversion controller. Desired discretetime dynamics for the subsystems was obtained by using Lyapunov synthesis method and bandwidth parameter selection method was also given. Simulation results demonstrate the rightness and validity.

The dynamic response of the pipeline system excited by the earthquake was modeled by finite element method and analyzed under consideration of the complex restrains of the pipeline system, including the pipe equipments, hangers, bends, etc. The 3D earthquake induced pipeline system vibration modes, the amplitudes of displacement and rotation responds, as well as the modal participation factor were calculated. The stochastic spectrum approach was used in the simulation of the earthquake response, applying the earthquake loads defined by the EIcentro spectra to the x,y,z directions over all of the pipelines. The research results showed that the earthquake has a great effect on the pipelines, it is important to take it into full account in the pipeline design, and the measures to enhance the antiearthquake capability were put forward.

Focusing on the problems of permeability measurement of inservice concrete structure, an indoor test system of relative permeability coefficient Ke of concrete was established, and electrical resistancepermeating time experiments for concrete were conducted. By Darcy's Law, the calculative formula of relative permeability coefficient Ke was deduced. In the meanwhile, according to permeating depth experiment rule(DL/T 51502001)， comparative tests of relative permeability coefficient Kr were completed. The research results indicate that depending on the descent characteristic of electrical resistance of concrete between the center hole and any observational hole in a permeating process, the permeating time can be measured, so that its anisotropy can be described and the value of relative permeability coefficient Ke of concrete was calculated. The experimental datafitting shows that the relationship between relative permeability coefficient Ke and Kr is a logarithmic function .

According to the mechanical structure and the load characteristics of a fighter airplane main undercarriage, the arrangement of the measuring strain gages and the type of the strain gage bridges were decided. Considering the variation of application of the loads and the influence of the landing gear compression stroke, an influence factor of landing gear compression stroke was defined and introduced in the calibration equation as a parameter. The equation between the load imput and the strain gage output was established on a test bench by aid of the multivariable regression and optimization method. The measured load spectrum of the main undercarriage was derived by means of the calibration equation, providing an important reference for estimating and enhancing the fatigue life and conducting the reliabilitybased design.

The trajectories of different diameter dp particle were simulated under the conditions of different jetcrossflow ratio R and jet exit diameter D's turbulent jettocrossflow, by using the renormalization group (RNG) kε model and SIMPLEC algorithm. The computation results showed that: ①It is approximately to symmetry in Y axis that the effect causing the particle by the flow field of round the jet hole distributes beside the XZ symmetry surface. With the increasing of the |Y/D|, the pull causing by the jet boost up at first and then take off. ②With the increasing of the jet diameter, the pull of the particle round the jet hole increase, and lengthen the journey in which the velocity of the particle is accelerated, the trajectory of the particle curved greater. ③With the increasing of R, the force of the particle suffer from the jet strengthen, ④The small diameter particle is good for following the flow, it's trajectories can reflect the trend of the flow field better.

The combined effects of temperature, pH, water/substrate ratio, enzyme concentration and hydrolysis time on the extraction of the soluble proteins, peptides and aminoacids from porcine hemoglobin by means of the enzymatic hydrolysis with the enzyme mixture of pancreatin and protamex were characterized by response surface analysis (RSA). The processing conditions were optimized by a mathematical model and the optimal parameter values were as follows: temperature is 49.4 ℃; pH is 7.7; water /substrate ratio is 1.4; enzyme concentration is 2.0 g/kg; and hydrolysis time is 18.6 h. The predicted nitrogen recovery(NR) value was 97.98%, and the actual value obtained was 96.94%. Four fractions of hydrolysate with average molecular mass >15 000 Da, 15 000\_5000 Da, 5000\_1000 Da and <1000 Da, were 21.1%, 3.9%, 20.9% and 54.1%, respectively. The hydrolysates had no bitter taste. The hydrophobicity data based on Q values support Ney's Q rule as a predictor of bitterness for porcine hemoglobin hydrolysates. So, the hydrolyzed proteins from porcine hemoglobin can be widely used in the food industry.