The problem of application of the driver model in the vehicle tire blow-out simulation and the changes of the vehicle response behavior after the tire blow-out were analyzed, and it was found that vehicle response changes after the tire blow-out are the main cause leading to the problem that the driver model can not drive the vehicle back to the original trajectory. A new driver model considering the fact that the vehicle is characterized by unsymmetric leftward and rightward turning after the tire blow-out was built and its simulation was performed. A vehicle state reference module was embedded in the original driver model and its parameters were identified under different driving speeds. Thus the problem of application of the driver model in the vehicle tire blow-out was solved, providing a reference for the study of driver operation behavior and stability control of the vehicle after tire blow-out.

In order to realize the accurate vehicle rollover warning, a 3-DoF rollover model and a time-to-rollover(TTR)-based rollover algorithm was established. Bionic intelligent artificial neural network(NN) was introduced to present a multi-parameter modified NN-TTR rollover warning algorithm. The typical vehicle state parameters were selected to analyze and correct the traditional TTR. The NN-TTR algorithm was validated using Matlab/Simulink and Carsim co-simulation platform. The results showed that the light vehicle rollover warning algorithm based on NN-TTR effectively improves the warning accuracy and provides a good foundation to improve the active anti-rollover control.

From the fact that the quasi-static method has deficiency in the solution of the transient mass center of sloshing fluid and the corresponding vehicle rollover modelling, the solutions of the transient mass center of sloshing fluid in tanks with modified square and modified oval cross sections were improved, and the exact equations for the mass center in the tank with modified square cross section were derived. A new vehicle rollover model was built without uncommon and hard-to-get parameters, simplifying the solution of vehicle rollover threshold. Using the improved quasi-static method, the change of the vehicle rollover threshold with filling percent under conditions of constant cargo density and constant cargo mass was simulated by Matlab. The results showed that the improved quasi-static method can predict accurately the vehicle rollover threshold.

A vehicle electronic stability control(ESC) hardware-in-loop(HiL) test bed and an ESC test car were established to form a platform for ESC system research and development.A hydraulic model was built for the ESC hydraulic control unit(HCU) and its parameters were calibrated according to the hydraulic characteristics acquired by tests on the ESC HiL test bed. Based on the acquired hydraulic characteristics of HCU, the open-loop estimation of the wheel cylinder pressure was performed. On the basis of this estimation, a closed-loop pressure estimation algorithm was downloaded to the ESC test car platform to do the ESC test. The test results showed that the proposed pressure estimation algorithm can provide accurate and reliable wheel cylinder pressure estimation values during braking force distribution for the vehicle with ESC system.

The structure of the ball-screw regenerative damper was redesigned according to the results of the previous research. Based on the velocity-force characteristic of the matched conventional damper, a 1/4 suspension simulation results were analyzed by the probability statistics method. Taking the enhancement of the energy regenerative efficiency as the target, the maximum power of the regenerative electromotor, speed and speed ratio were matched. The ball-screw parameters were matched and verified in the aspects of structure and efficiency. The results showed that the designed and matched regenerative damper can effectively regenerate the suspension vibration energy, satisfying the actual requirements to replace the conventional damper.

By analyzing the relationship between the gearbox output torque and the resistance torque in the classic dynamics equation of the vehicle, the traditional load concept was extended to the generalized load, and a torque-based vehicle load recognition method was proposed to detect the informations of the road slope, the cargo magnitude and the road surface resistance during the driving of the vehicle with automatic transmission. Requiring only the common sensors, the proposed method recognizes the extended or generalized load from the difference between the current gearbox output torque and the resistance torque under zero-load condition. The principle of the load recognition was described in detail, the effects of different driving conditions on the load recognition were analyzed, and the detailed load recognition steps were given. The vehicle tests showed that the method is practical and reliable, can effectively recognize the load informations during the driving of the vehicle with automatic transmission.

In order to decide the rational wheel number of the lunar rover, a wheel-soil contact model for the wheel at moving forward and turning was built based on the theory of terrain mechanics and the consideration of the sand springback sinkage. The calculation expressions for total resistance torques of the single wheel at moving forward and turning were derived. The relationship curves of the driving torque, the resistance torque at moving forward and the resistance torque at turning versus sinkage were achieved under a certain range of wheel width and radius. The sinkage interval to make the wheel have high driving efficiency was found by analysis. Based on these results, the flow chart for choosing wheel number rationally was proposed, and the advantages and disadvantages of the rovers with different number of wheels were compared to provide a reference for the comprehensive and rational wheel number choosing.

A formula for the interior sound pressure of the rectangular rigid box with a flexible wall and a new method to calculate the mode shape coupling coefficient by the finite element analysis were derived and validated. The particular points of the mode shape coupling coefficient were analyzed in detail, and it was found that the main factors affecting the coefficient are the mode shape similarity and complexity of the structural sound cavity in coupling surfaces which affect the coupling degree between the structure and acoustic field. The results were used to analyze and control the interior noise of a car, and the noise reduction of 5 dB was achieved at the driver right ear measuring point, providing an effective new method for low interior noise design of the car body.

In order to deal with the problem that the tracked vehicle with hydraulic drive steers the pressure of the outer hydraulic system may become too high and the problem of the steering trajectory control, a new steering control strategy was proposed. In the proposed strategy the outer side hydromotor displacement uses dual-parameter control by hydraulic pressure and engine speed, while the inner side uses speed closed-loop control to follow the outer side. The targeted relative steering radius, the turning angle of steering wheel, vehicle speed, hydraulic displacement, engine speed, and hydraulic system pressure were matched. A simulation model was built for the whole vehicle steering, and a test system was established for the real vehicle. The results of simulation and field test showed, the proposed control strategy can control effectively the hydraulic system pressure, the parameter match can meet the requirements of steering control, to realize the accurate and rapid steering of the vehicle according to the targeted relative steering radius.

To optimize the blade forward-curved angle of the hydraulic retarder, a 3-D integrated optimization simulation platform was built. Taking the blade angle as the design variable, the design of experiment was performed using the CFD analysis as the solver, and a response surface method model was built according to the test sample results, the optimization was accomplished. A set of the optimized blade curved angle parameters of the rotor and stator wheels were obtained. The effects of the blade angle parameters on the internal flow characteristics and the brake performances were analyzed, and the brake performances before and after the optimization were compared using the test data. The results showed that the brake performances of the hydraulic retarder with optimal blade angle parameters were improved obviously. The built 3-D integrated optimization platform is accurate and reliable.

By the characteristic analysis of the typical working conditions of the hydraulic excavator, a boom energy-saving system with closed oil circuit was proposed and the energy-saving principle of the system with the energy accumulator was discussed. The mathematical models were built for the hydraulic and the electric components according to their working principles. The working conditions of its 3 main power components, the accumulator, the electromotor, and the load in the typical excavator working cycle were analyzed. The results of the theoretical analysis and the simulation showed that the boom energy-saving system with closed oil circuit works steadily. Compared with the traditional valve-controlled system of the hydraulic excavator, the energy-saving rate of the proposed system reached, providing an encouraging reference for the energy-saving of the hydraulic excavator.

The shear yield property of electrorheological (ER) polishing fluid was investigated for ER fluid-assisted polishing. The microstructure of fibrous columns formed by particles perpendicular to the electrodes was observed by electronic microscope when the electrical field was applied. A theoretical model for the static shear yield stress of the ER polishing fluid was built to calculate the shear strength of the chain structure of particles, which was based on the combination structure of ER particles with abrasive particles and the attraction forces between particles in ER polishing fluid. A series of experiments were conducted to measure the static shear yield stress of ER polishing fluid by the developed yield stress testing device. The effects of the species, sizes and concentrations of abrasive particles on the shear yield stress of ER polishing fluid were obtained and the validity of the presented model was confirmed by experiments.

A new way to calculate the instantaneous availability of the computer numerical control(CNC) machine tool was proposed based on the functional data analysis method. Using the quartic Bernstein polynomial as basis function to smooth the reliability estimation values, as well as the least square method to estimate the fitting parameters of the basis function. The estimated maintainability was across-section adjusted according to the trouble occurrence time section. Based on the adjusted trend of maintainability, the Fourier basis function was chosen to smooth it. The fitting coefficients were solved by the non-linear least variance method. A model was built for the instantaneous availability of the CNC machine tool based on the functional linear regression model, and taking 18 machine centers as examples to verify the proposed model. The results showed that the proposed model gave out the correlation coefficient of 0.9641, being effective to evaluate the instantaneous availability of the CNC machine tool.

Based on the most popular standard called MIL-STD-1629A from the U.S. Department of Defense, an improved FMECA method was presented to investigate the CNC turret. To avoid the ambiguity of subjective evaluation, the importance degree was used to judge severity, weighting factors of experts were introduced to determinate the criticality, and a fuzzy logic-based expert system was adopted to evaluate the risk level comprehensively. The weak link affecting turret system reliability was found out, which provides a scientific basis for improved design and maintenance policy.

The surface characteristics of the wings sparrow hawk and long-eared owl were extracted and used to optimize the blade of small axial fan to improve aerodynamic performance of the fan. The best fan was selected from eight bionic axial fans using CFD method. Computer simulation analysis and aerodynamic experiments show that the flow-static pressure curve of the bionic blade fan is remarkable better than that of prototype fan. The maximum mass flow rate of the bionic fan is 6.1% higher than that of prototype fan and the maximum static pressure is 7% higher, respectively. Experiments also show that the V-section is better than arc-section of the blade. The study provides a preliminary basis for optimizing the parameters of the blade of bionic fan.

The frosting behavior in the cooling process of the copper flake specimen as cold surface was observed microscopically. In order to avoid the obstacle of the frost layer growth at the specimen edge to the observing area, the specimen was bent to an arch to observe the frosting phenomena at the arch apex. The experiments showed that the initial rate of the frost layer formation is different from the subsequent growth rate. In the frost layer growth process, in addition to the melt-back phenomenon of the surface frost branches, there were periodic overall collapse of the outer frost layer. Analysis suggested that the breeding point distribution and density of the initial frost layer are the origins of difference of the frost layer growth rates, and the collapse of the frost layer is the result of the melt-back and recomposition of the initial loose frost layer.

Adding small amount of Fe-based additive into the low-sulfur (w(S)=30?10^-6) diesel fuel to investigate experimentally the effects of Fe on the diesel engine power, fuel economy, gaseous emissions and particulate emissions including particle number concentration and size distribution. Results showed that Fe addition leads to a slight increase in power output, a slight decrease in specific fuel consumption, and more or less decrease of CO,HC,and NO_x emissions. Fe additive produces the additional nano-particles beyond the diesel particulates and increases the total particle number concentration, especially the nucleation mode particle. Under all test conditions, Fe additive makes the ratios of the nucleation mode particle higher than 90%, and always higher than those using the standard diesel fuel without Fe additives. Use of Fe additive leads to an obvious reduction of the accumulation mode particle emission except under high-speed conditions.

For the underground energy storage system in the region with strong groundwater seepage, a new offset load control mode was proposed based on a large-scale numerical model of multi-heat source. Applying the dual control of the load magnitude and the temperature potential, it was possible to enhance the heat transfer efficiency of the underground energy storage system and the energy consevation capability of the energy storage region. The analysis demonstrated that at the condition of same energy input, the offset step-load control mode is superior to the traditional ring-load control mode in the capacity of energy storage and conservation. The study on the offset dynamic energy storage intermittence showed, short intermittence is superior to long intermittence in the energy storage magnitude, the average temperature rise and the maximal temperature to enhance the energy storage, diffusion and conservation in the pipe burying region.

In order to alleviate the road network jam, a balanced control method was proposed for the road network state. The road intersection saturation degree was chosen as the traffic state index, a bi-level programming optimization model was built for the balanced control of the road network traffic state to minimize the maximal road network saturation degree. This optimization model was solved by the genetic algorithm to optimize the traffic signal cycle and its green-red split. The balanced control was improved by consideration of the delay index, traffic state, and the relative importance of each intersection. The numerical simulation results showed that the proposed method can effectively alleviate the traffic jam in the critical region.

For the improvement of the effect of incident detection using the single automatic incident detection algorithm (AIDA), a set of indexes used for traffic condition on-line evaluating was designed based on the analysis of the causes of failed alarms and false alarms generated by AIDAs used, and an AIDAs fusion method based on factor analysis and cluster analysis was developed. The proposed method was verified and compared using the data collected from the inductive loops on a metropolitan urban freeway. The results showed that under the condition of aimed false alarm rate 0.5% and the detection rates of the original AIDA 63.5 %~66.1%, the detection rate of proposed method is 90.6% and its false alarm rate is only 0.0981%, being obviously better than the contrast methods.

The car-following behavior was studied by molecular kinetics, the required safe distance was explored and an application model was built. By giving the reaction weights to various stimuli, a state equation for the isodirectionality of the responses of the driver to various stimuli was derived using the mathematical deduction. Based on this equation, a molecular car-following model was built to describe the car-following behavior. The results of analyses from the data validation and the experimental comparison showed that the proposed molecular car-following model describes the car-following behavior more completely and appropriately, makes up the deficiency of the speed-based car-following models, and improves the car-following theory.

Microstructures and mechanical properties of Nd: YAG laser welded transformation induced plasticity (TRIP) steel with tensile strength 650 MPa and 1.2 mm thickness were studied. The hardness and tensile strength of butt welding joint were investigated by optical and electronic microscopies under different welding speeds. The formability of raw material and welding joint were tested with Erichsen experiment. The study showed that the welded metal mainly consists of martensite which is different from the original TRIP or parent metal, which is composed of ferrite matrix plus bainite and a little residual austenite. Therefore, the welded metal or the heat affected zone has the maximum hardness for the whole welded joint. The yield strength (YS) and tensile strength (TS) of the welded specimens tested perpendicular to the weld line were almost equal to those of the parent metal. But the YS and TS of the welded specimens tested parallel with the weld line were obviously higher than those of the parent metal. The stamping formability of laser welded TRIP steel sheet was obviously decreased compared with that of the parent metal.

A multi-point forming with wrinkle resistance function(MPFWR) was introduced which doesn't need the edge-pressing ring. This forming mode is different from the common multi-point die forming(MPDF) in that the winkle resistant holders and form adjusting holders arrange alternately. Holding the whole blank in the forming zone, the discrete wrinkle resistant holders can float up and down along with the blank. The forming process of a spherical part in MPFWR was simulated using a finite element analysis software. The effects of the sheet metal blank thickness and the wrinkle resistant holder force on the maximum deformability were analyzed, and the results were compared with those of MPDF. It was found that the MPFWR can restrain the wrinkle of the sheet blank, gives out the obviously larger formability of the spherical part. The thinner of the blank, the better its formability. The wrinkle resistance can be enhanced by increasing the holder force.

The topological relations between a region with two holes and a simple region were studied. First, we extend 4-intersection matrix model to 16-intersection matrix model to represent the spatial relations of four simple regions, a region with two holes can be seen as a simple region A and two simple regions B and C included in A. Then, a region with two holes and simple region can be seen as four simple regions satisfying certain restricted conditions. Consequently, we obtain 70 topological relations between region with two holes and a simple region in practice by programs and we prove that these topological relations are exclusive and complete. Moreover, compared to the topological relations given by RCC-5 in pairs, the 16-intersection matrix model has stronger expressive ability.

On the basis of the research of the constraint propagation algorithm, an adaptive constraint propagation algorithm, AC_MaxRPC_Bitwise, is proposed, which is based on bitwise operations. The proposed algorithm has several advantages in improving the efficiency of adaptive constraint propagation method. First, it introduces bitwise method to represent the data structure while looking for AC support and PC support. Second, it uses bitwise operation to speed up the searching process of AC support and PC witness. Experiments were conducted on a few typical benchmark problems. Results show that the improved algorithm AC_MaxRPC_Bitwise whelms AC and other constraint propagation algorithm in overall performance.

Simulated Annealing Algorithm (SAA) is used to dynamically adjust the parameters of Dynamic Fuzzy Neural Network (SAA-DFNN). The SAA-DFNN is applied to study Software Reliability Growth Model (SRGM). The SAA is used to resolve the optimal solution of DFNN parameters in the DFNN software failure data training process. Then according to the obtained DFNN optimal parameters SAA sets up software failure data prediction model. Using three groups of software defect data, the predictive ability of the SRGM established by SAA-DFNN is compared with that of the SRGM established by Fuzzy Neural Network (FNN) and by BP Neural Network (BPN)and G-O model. Simulation results confirm that the SRGM established by SAA-DFNN has steady single-step ahead predictive ability with certain versatility and the prediction error is smaller.

The Fault Induction Attack (FIA) proposed by Adi Shamir is an effective method in cracking RC4 algorithm. The target of the FIA is the Pseudo-Random Generation (PGR) phase of the RC4 algorithm. To resist this attack, it is necessary to enhance the security of the PGR phase. This paper presents two improvements that make the nonlinear transformation more complex by adding the row shift and column shift to the PGR phase. After analyzing the validity, security and the efficiency of the FIA, it is demonstrated that the efficiency of using FIA is lower than using exhaustive attack, which means that the improvements could resist FIA. Comparison of the two improvements shows that the first improvement is more efficient and the second improvement is more secure.

An energy-efficiency amendatory algorithm is proposed to save limited energy in Wireless Sensor Network (WSN). In this algorithm, a sensor node changes it status according to the threshold of its energy condition. Therefore, the energy consumption in the path with the sensor node consuming most energy could be transferred to other paths in order to reduce the whole energy consumption of the network. Through experiments and analyses the relationship curve between the probability of number of nodes and the amount of data transmission was obtained; the factors that affect this relationship curve were analyzed; the features of the curve and the threshold mentioned above were found. Because of the energy consumption by the algorithm itself, the concept of data gravitation was proposed to reduce the energy consumption. Experiment results show that the algorithm is energy efficiency and the lifecycle of the networks is increased.

A real-time system must guarantee the stringent requirements of real-time response and reliability. Taking account of the effects of errors, an investigation of the schedulability of fault-tolerant real-time systems was conducted based on the rollback recovery fault-tolerant model. A priority mixed strategy of fault-tolerance was proposed, and the computing formula of worst-case response time of a task under the new strategy is deduced. The new strategy allows the faulty task to execute at higher or lower priority levels. According to the schedulability analysis, an efficient priority assignment search algorithm (FTPCS_MS) was proposed. The algorithm reduces the search space of optimal fault-tolerant priority mixed assignment from O(nn) to O(n²). The simulation shows that the mixed strategy can improve system fault resilience efficiently.

A new optimized detector generation algorithm is proposed to overcome the shortcomings of available real-value variable-radius detector generation algorithms. By statistic analysis of the detector generation, a hypothesis testing based detector generation process is proposed. The result of the hypothesis testing is taken as one of the control parameters to end the algorithm, thus, it can effectively avoid the generation of redundant detectors. Meanwhile, the algorithm makes full use of the distribution of self-space, optimizes the center position and expands the radius of the detectors in order to generate the detector with large coverage. The 2DSyntheticData, the actual Irish data set and biomedical data set are used to test the algorithm. Experiment results show that the algorithm performs very well that it improves the detection rate, reduces the number of required detectors.

Two Format-Preserving Encryption (FPE) schemes suitable for variable-length encoding character data were proposed. Scheme I constructs the block cipher based on Feistel network, and combines cycle-walking to ensure that the ciphertext is in legitimate range. This scheme can be applied in FPE on short plaintext, and meets the needs of hiding the count of characters. Scheme II splits the plaintext into several sub-character sets, and then it executes the FPE of fixed-length encoding character data on each sub-set. Scheme II is more efficient and applicable to FPE on plaintext of any length.

Traditional methods of Differentially Expressed Genes (DEGs) analysis can not be used to deal with heterogeneous data sets that the analysis results are inconsistent usually. A new method, named Rank Standard Deviation Meta (RSDM) analysis, is proposed in this paper for detecting DEGs. The method is based on the meta-analysis and rank standard deviation filtering technology. The proposed method can detect True Differentially Expressed Genes (TDEGs) and filter Pseudo Differentially Expressed Genes (PDEGs), both TDEGs and PDEGs coming from experimental datasets. The experiment results show that the propose method is of high efficiency.

Based on machine vision, an intelligent counting system on soft fibrils collection was designed. The hemispherical LED source was used in the system, and the fiber slice was manufactured. The fiber slice was imaged by a microscope optical system, recorded by a CCD camera, and transferred by the image grabber into a computer. Then the quantities of the soft fibrils collection were calculated by the image processing system, and whether the numbers were qualified or not could be judged. A novel image processing method was proposed. The original images obtained in different light intensity could be fused based on maximum region entropy. The optimal threshold of the fusion image could be got based on SOFM neural networks. Based on the above threshold, the binary images of the originals could be refused. After that the quantities of the soft fibrils collection could be counted through boundary separation and counting algorithms based on statistical value. The experimental results show that the detection error of the system is less than 1%, the maximum standard deviation is no more than 0.07 and the automatic and intelligent counting function on the soft fibrils collection can be accomplished.

To solve the "false event" detection problem in image sequence analysis in the wild situation with strong background noises, an event detection method based on temporal-frequency denoising is proposed utilizing Kalman filter theory and 2D Discrete Cosine Transform (DCT) theory. The method utilizes Kalman filter to denoise the background model of image sequence with several frames in time domain. The background model is reconstructed adaptively based on variational foreground region. Then, the foreground image of a single frame is denoised utilizing 2D DCT transform. Finally, the event foreground is segmented by adaptive segmentation method. Based on the results of simulation analysis of image sequences collected in wild situation, the proposed method is proved to detect the event foreground and solve the "false event" detection problem effectively and practically, while its F-measure can achieve 0.9423.

To accomplish the synchronization of nonsinusoidal orthogonal modulated signals, an external synchronization method based on double Phase-Locked Loops (PLLs) is proposed. An auxiliary sequence made up of bandpass Prolate Spheroidal Wave Function (PSWF) pulses is used in the method. The auxiliary sequence is separated from the modulated signals in the frequency domain. In the synchronization procedure, the carrier synchronization is first achieved by the narrowband PLLs. Then on this basis the baseband PSWF sequence is obtained by the coherent demodulation. Finally, the digital pulse PLLs is employed to achieve the symbol synchronization. Simulation results show that the proposed method can accomplish the synchronization quickly, and it has good robustness and strong ability to overcome noise. In particular, the method is applicable to the system, which has many parallel modulated subbands.

This paper first introduces the optimization method on node transmission power based Quality of Service (QoS), and then discusses the node power consumption model. The power consumption is taken into account for the calculation of air-time metric, which is also regarded as an important factor in path selection. The improved power consumption model and packet transmission delay are simulated as the output in NS-3 network environment. It is shown that the improved method effectively reduces the node power consumption, while the introduction of power consumption parameter does not bring too much delay, thus, the user QoS requirement is ensured.

A clock synchronization algorithm in wireless measurement and control network is proposed. It is based on the square-root Kalman Filter to achieve the optimal estimation of the clock offset and skew on the premise of algorithm convergence. The algorithm also puts forward a model of communication cost and clock accuracy to calibrate the clock under certain clock accuracy. This can ensure the stability and accuracy of the clock and reduce the communication cost. Some theoretical and simulation results are presented to illustrate the performance of the proposed clock synchronization algorithm.

According to the actual need of field geological exploration, a wireless network monitoring technology is proposed based on low-power WiFi broadband wireless communications. It is used to monitor the front-end state information and the quality of collected data by setting up field WLAN. It provides low-cost, low-power and high reliability communication means and wireless monitoring method for high-density, three-dimensional and large number channels deep exploration. The cable-less stored seismograph can be measured and controlled by wireless method. It breaks through the defect of sealing property and improves the flexibility and reliability of the cable-less stored seismograph.

A grid resource Management model based on Combinatorial Double Auction with Virtual Seller (MCDA+VS) is proposed. In MCDA+VS mechanism, a dynamic bidding strategy is established to afford participants opportunities of re-bidding in different round of an auction. As a result, the transaction ratio can be effectively increased. In addition, a virtual seller is introduced and involved in each round of auctions to transform surplus resources into trade revenue. The idea of utilizing surplus resources can lead to not only higher resource utilization but also larger trade revenue. In grid environment, the improved auction performances help to establish good economic incentive, which could attract more grid users and resource providers into grid transactions.

A miniaturized dual-band Wilkinson power divider with a coupled-line section is presented. The output ports are isolated physically to reduce parasitic effects and to make an interconnection with other passive-active devices easy. Closed-form design equations based on an even-and odd-mode method are obtained. The design methods are discussed in details. The measured results of a microstrip power divider operation at 1 GHz and 2 GHz verified the theoretical feasibility of the proposed power divider.

Existing studies of the link dynamics in Mobile Ad hoc Networks (MANETs) are mainly based on some traditional mobility models, such as Random Direction (RD) model and Random Waypoint (RWP) model etc. For these models, the unrealistic moving behaviors, such as sudden speed change and sharp turn, may invalidate the analytical and simulation results of link dynamics. The 3-D Smooth Gauss-semi-Markov (SGM) mobility model, which can truly reflect the physical law of moving objects in 3-D space, and the Markov chain theory were used to study the link dynamics in 3-D MANET. The analytical expressions for the link breakage probability and the expected link lifetime were derived from the probability density function of the relative velocity and the distance transition probability matrix between two nodes. Finally, the derived expressions were confirmed through simulations. The conclusion is valuable for the design of network protocols in 3-D MANETs.

To solve the problem of audio-video synchronous coding effectively, an embedding scheme is proposed. This scheme applies the quarter-pixel motion search during the motion estimation in H.264. At the transmitter terminal, according to the mapping rule between the binary audio bits and the search points, the audio bits are embedded into the video signal by modulating the best search point in the quarter-pixel motion estimation. Then the synchronous coding is applied to the hybrid signals. At the receiver terminal, the audio information bits are extracted from the video stream according to the embedding rule. Then the audio and video signals are reconstructed respectively. Experiment results show that the proposed scheme can achieve audio-video synchronous coding with low embedding cost and good performance in quality of video and audio signals.

To solve the problem of oversimplified speech short-time spectral amplitude distribution model in speech signal processing, an estimation method of speech amplitude spectrum based on hidden Markov model is proposed. This estimation method uses Rayleigh mixture model as speech amplitude spectrum distribution and employs hidden Markov model to divide speech signal into different states. In each state, there is accordingly a group of Rayleigh mixture model parameters. By the division of speech signal into different states, this method can achieve speech classification and build more accurate model for speech signal short term spectrum.

To improve the accuracy of Voice Activity Detection (VAD) under noisy Environment, a voice activity detection algorithm based on Mel Frequency Cepstrum Coefficient (MFCC) similarity is proposed. First, the MFCC from each frame of the voice signals is extracted. Then, the first ten frames are taken as the background noises. Finally, by calculating the MFCC correlation coefficient distance under the above noisy condition, the voice-activity parameters are detected from the MFCC similarity curves. The experiment results show that the proposed algorithm is effective under both white noise and pink noise conditions and at low signal to noise ratio.

Egg yolk powder was obtained by spray-drying technology, and the powder properties were studied. On the basis of single factor test, the effects of the factors, such as spray flux, feed temperature and inlet air temperature, on the solubility of the egg yolk powder were investigated by Box-Benhnken center composite design and response surface technology. The powder properties were analyzed through its rheological property, granular structure and stability. The results show that the optimum technical conditions are as follows: spray flux of 17.37 ml/min, feed temperature of 35.65 ℃, inlet air temperature of 185.36 ℃. Under the optimum conditions, the solubility exponent is 1.8506 and the solubility is 98.74 g/100 g. Analyses show that the instant egg yolk powder is typical non-Newtonian fluid but with pseudo plastic characteristics. Moreover, the powder has intact structure and stability. Compared with fresh egg the recoverability of the egg yolk powder is better.

The influences of the ratio of xanthan gum (XG) to sodium carboxymethyl cellulose (CMC) mixture and the pH on the behavior of oil-in-water caseinate emulsion were investigated. Zeta-potential, droplet size, viscosity, creaming behavior and the microstructure of the emulsion were detected to reveal the possible mechanisms of the stability or instability of the emulsion. Results show that, under neutral condition, both anionic polysaccharides were not absorbed to the surface of the caseinate-coated droplet, which caused the depletion flocculation. The creaming behavior of visual separation was quickly observed during the storage time. Under acid condition, xanthan gum or sodium carboxymethyl cellulose was absorbed to the droplet surface and increased both the electrostatic and steric repulsion of the emulsion. At certain polysaccharide concentrations (CMC∶XG = 1∶1, 3∶1,the total addition amount of the emulsion is 0.4 wt%), the stability of the emulsion was enhanced, and during two weeks room temperature storage there was no notable creaming phenomenon.