Based on the infinite impulse response digital filter theory,a frequency weighting filter was designed according to the frequency weighting factors in the standard ISO2631-1:1997 (E).The weighted acceleration data in the timedomain and the crest factor of vibration were obtained by the filtering of the time-domain acceleration signal using the fuequency weighted digital filter. The analyses showed that the crest factor of the vibration in all axes is lower than 9 under random vibration, the basic evaluation method can be used to evaluate the vehicle ride comfort. Under the conditions of occasional shock and transient vibration, the crest factors in the vertical direction of the seat, the seat rail and the under-seat floor will be higher than 9, and the additional evaluation method should be used. The results showed that the vehicle ride comfort weighted evaluation in time-domain can be achieved by the frequency weighted digital filtering, it can compensate the inadequacy of the traditional weighted evaluation method in frequency domain in being unable to get the crest factor, and enhance the accuracy of vehicle ride comfort evaluation.

The qualitative and quantitative evaluations of the automobile comprehensive performance were performed using the radar chart comprehensive evaluation method. The dynamic performance, the fuel economy, the driving safety, the handling stability and the ride comfort were chosen as the evaluation indexes, a reasonable evaluation system was established. The evaluation indexes of 4 automobiles were normalized to eliminate the impacts of their dimensions and dimensional units. The comprehensive performances of the 4 automobiles were evaluated using method of improved radar chart which is characterized by the result uniqueness. The weights of the evaluated performances were anlyzed by the analytic hierarchy process to make the evaluated results more comprehensive and reasonable.

The method for motors matching was put forward with the help of the analysis on the characteristics of the dualmode hybrid transmission. The motors' parameter matching was studied on the basis of the analysis on the characteristics of the power split and the relations of the two motors' speed and torque and those of the system. As a result, the relations among the parameters of the engine, the output characteristics and the structural parameters of the dual-mode hybrid transmission and the parameters of the motors were revealed.

To deal with the fact that there are uncertainties such as disturbance signals, sensor noises,etc., in the electric power steering(EPS) system, and the demand of the system to the robustness and the dynamic performance, a road feel model was built for the EPS system and an H_∞ controller was designed for the EPS road feel. The output performance of the system was shaped by selection of the frequency weighting function, and a simulation was performed in the frequency domain. The simulation results showed that compared with the traditional PID controller, the H_∞ controller for the EPS road feel can provide a better system robustness, the frequency weighted EPS can acquire better road informations at low frequencies, can attenuate the noises and disturbances from the road at high frequencies, thus insure the driver to get a better road feel.

The forming simulation, the experimental test, and the structure analysis in combination were applied to study the sheet thinning and the material work hardening in the stamping process of the car body component. It was found that in the forming process, the yield limit increase of the sheet metal demonstrates a more significant effect on the enhancement of the car body component strength than that of sheet thinning on the weakening of the strength. The obtained results provide a more accurate reference for the optimization on the light-weight design of the car body structure.

In order to improve the simulation accuracy of the weld spot of the front longitudinal beam in the vehicle crash simulation, a spot weld model using the shell element to simulate the parent material and the single solid element to simulate the weld nugget and its heat affected zone was built by means of the software LSDYNA. The properties of the parent material and the failure parameters of the weld spot were acquired by tensile tests of the sheet and the weld spot cup respectively. The internal force of the spot weld model was investigated by the numerical KSII tests in various parent material mesh configurations. The simulations of the front longitudinal beam crash were performed by a validated finite element model of the longitudinal beam using spot weld models with sodid element and beam element respectively. The simulation results were compared with the sled test results in terms of the failure behavior of the spot weld, the deformation of the longitudinal beam and the acceleration of the sled. The study indicated that the spot weld model with solid element is better than that with beam element for better reliability and higher accuracy.

The comparisons of fuel economy and emissions between the series plug-in hybrid electric vehicle(PHEV) and the equivalent prototype vehicle were simulated using the CRUISE code under different driving cycles, such as urban driving cycle(UDC), new European driving cycle(NEDC), extra-urban driving cycle(EUDC), and Chinese city driving cycles on expressway and common road for passenger car, to investigate the energy-saving and emission reducing effects of series PHEV. The results showed that the minimum fuel consumption of PHEV in different driving cycles is about 2.1 L/100 km which is 62% less than that of the prototype vehicle. At meantime, the series PHEV emits significantly less emissions, its emissions of HC,CO and NO_x is reduced by 61%,48% and 41%, respectively by comparison with the corresponding Chinese emissions standard(IV).

The effects of the fuel injection pulse width, the ignition advance angle and the parameters of engine start PID controller on the engine starting behavior was researched on a turbocharged gasoline direct injection engine with the selfdeveloped software and hardware of the engine control system. The test results showed that the fluctuation of the engine starting speed is small when the fuel injection pulse width is  45 ℃A. The engine starting torque is the maximum for the ignition advance angle 12 ℃A. The optimal parameters of the PID control system are as follows: the speed responses quickly at  K_P=1; the errorfree speed tracking and the small overshoot are reached at K_I=0.2; and the control time is short and the speed becomes steady quickly at K_D=1.

An experimental investigation is conducted to evaluate the effects of using GTL fuel on the particle size distribution of a fully instrumented, turbocharged and intercooled, common-rail direct injection (CRDI) engine. The tests are conducted using diesel , GTL fuel and their blends, with the engine working over three transient cycles. The results show that the total particulate emission and their growth with the elapsed time reduce as the transient rate reduced during the three transient cycles. The total particles, nuclei mode particles and ultrafine particles have all reduced as the increase of GTL fuel to the blends, but it increases a small margin for the accumulation mode particles. The nuclei modes particles for G20 and G30 fuel decreased 42% and 20%, respectively, compared to the diesel fuel. The percentage of nuclei modes particles and ultrafine particles for G20 and G30 fuel have a conspicuous decrease, and it increase a small margin for the G10 fuel during all the transient cycles.

The control patterns of loading period in the field of underground energy storage were investigated based on a large scale multi-heat source model, and the intermittent sequence and dynamic storage modes were considered accordingly. The results indicate that the storage efficient in the period of heating is affected by the dynamic temperature field significantly which is formed by the temperature amplitude distribution, intermittent period, rest cycle in the different load distribution, while the energy diffusion in the period of energy reservation is also affected by the dynamic temperature. The controllable energy diffusion transfer method both facilities energy storage and inhibits the loss of the stored energy. The results also indicate that the semi-day intermittent mode is more effective than the full-day intermittent mode, and the shorter intermittent period is more beneficial to optimize the energy storage quantity, average temperature and maximum temperature during the energy storage. Therefore,a reasonable selection of dynamic load mode with intermittent period and temperature distribution is beneficial for efficient heat transfer and effective energy transportation in multi-heat sources system.

A model was built for the tractor-semitrailer in the road reference frame to analyze the response of the tractorsemitrailer to the tractor lateral displacement. The robust control theory and algorithm were used to study the disturbance control of the tractor-semitrailer. A H_∞ loop-shaping controller was designed by techniques such as the pre- and/or post-weighting of the nominal controlled object and the normalized left coprime faetorization. The designed controller appears to be characterized by a good trade-off between the driving performance and the robust stability, ensuring the driving performance of the tractor-semitrailer under the lateral force disturbance with little lateral track shift.

Taking the collision avoidance in the pedestrian simulation as the research object, a concept of blockade-angle was introduced. The blockade area that needs to circumambulate by the pedestrian was defined by the introduced blockade-angle of the impediment to the on-going pedestrian. The circumambulation route of the pedestrian around the static impediment was chosen based on the shortest route principle. Considering the right-hand walking custom in China, the colliscon avoidance of the counter-directional pedestrians was realized by reducing the priority of the end-side of the blockdae-angle. It was suggested that for the special case of a pedestrian being limited by all sides of the blockade-angle, the question that whether the pedestrian should move continually and how much is his speed was depended on the speed and direction of the nearest pedestrian on the target direction. The realizing process of the proposed collision avoidance method was simulated, the results showed that this method does not require the complicated calibration process, the modeling process in the pedestrian simulation is also greatly simplified.

An MPI+OpenMP hybrid programming model was proposed based on the Floyd algorithm as its core to solve the shortest path problem which needs huge computation and high real-timeness. The proposed model provides the intra- and inter-node hierarchical parallel processing, takes full advantage of shared memory model and message passing model, improves the system performance, and enhances the calculation speed. The model was validated by the road network in Changchun city, and the results showed that the proposed model is characterized by better speedup ratio and operation efficiency than the MPI model. With the increase of the node, the speedup ratio of the model improves more significantly, indicating the model has a better expansibility.

To describe the process of queue growth and dissipation on the link between coordinated signalized intersections, a vehicle queue model based on the shock wave theory was developed by analyzing the evolution of vehicle queuing. In order to validate the reliability of this model, a simulation scheme was designed by using VISSIM traffic simulation software to collect the speed and arrival time of each vehicle at upstream and downstream detectors. The queue dissipation lengths and residual queue lengths for 55 cycles were calculated from the proposed model. The outcomes reveal that the cycles in which the relative errors between the simulated and calculated values of queue dissipation length and residual queue length are less than 10% account for 72.73% and 85.45% respectively. This model can be applied to estimate the queue length on the link between coordinated signalized intersections, providing a theoretical basis for optimization or adjustment of traffic control scheme.

A new 6-DoF state ego-motion estimation algorithm was presented based on the lunar rover stereovision image sequence to realize the autonomous navigation of the rover. A scale invariant feature transform SIFT feature tracking and matching method was proposed and studied. 2 new matching and restraint conditions were introduced to establish the accurate environment feature correspondence before and after the motion of the rover. A robust estimation technique was proposed to solve the motion parameters of the rover based on the obtained feature correspondence. A method combining random sampling consensus and unit quaternions was used to reject the matching error and get the preliminary estimation of the rover body motion parameters, and from this preliminary estimation, a Levenberg-Marquardt non-linear optimization algorithm was used to find the accurate solution of the motion parameters. The effectiveness of the proposed motion parameter estimation algorithm was verified by the simulated field test.

The spatial motion pattern recognition method and the abnormal traffic hehavior detection method of the moving object were proposed. The spatial distance similarity matrix of the trajectory dataset was constructed using the mearurement method in the trajectory space with the improved Hausdorff distance. The typical motion pattern of the moving object was extracted by the spatial distributions of the trajectory dataset learnt by the spectral clustering algorithm. A motion pattern matching algorithm in the trajectory space was presented for the abnormal traffic behavior detection based on Bayes classifier.The vehicle lane-changing behavior in the realistic scenario was recognized as an example, and the result was encouraging. The proposed method provides a technical support for the studies of the traffic behavior of moving object and the intelligent traffic management and control.

Taking the time delay spread between the successive flights at the same gate as the measurement of the airport service quality level, an optimization model for the airport gate assignment was built. The model provides the airport controller with an optimal airport gate assignment to reduce the coming influence of the delay as much as possible. The model was solved by the tabu search algorithm using the practical data from a certain airport in China. Simulation results showed the validity of the proposed model.

The weight vector of the evaluation indexes was determined by the fuzzy hierarchy analysis, the membership functions of the evaluation indexes to the evaluation class were established using the ridge function, and a BP neural network model for evaluation of the vehicle safety status was built using the neural network tool box in Matlab software. The experiments were performed using the preset fault test method, the operation parameters of the vehicle under the road test process were acquired by the vehicle monitoring and early-warning system. The training samples for the neural network were obtained by the fuzzy comprehensive evaluation, and the trained BP neural network was used to evaluate the vehicle safety status in the road test. The road test results showed that the neural network evaluation is more sensitive and more accurate than the fuzzy comprehensive evaluation, providing a new method for identification of the vehicle risk status.

Residential location and commute choice are jointly related decisions among station-area residents. The impact of rail transit on residential location and commute mode choice was analyzed from a microscopic perspective. On account of the advantages of cross-nested logit model structure to allow capturing the potential spatial correlation between spatially contiguous alternatives, cross-nested logit model was specified to estimate the joint residential location and mode choice around the rail transit stations. Factors as not only the individual and socio-economic characteristics but travel and land use attribute as well were defined as exogenous variables. This model has been calibrated in biogeme. In addition, the change of alternatives probability brought by factors variation was simulated. The results show that those living nearer rail transit stations tend to rail-commute far more often and when the car commute time increase the nearer residents live rail transit station, the higher probability rail-commute shift.

A study on the abrasive wear behaviors of pure Mg and (Mg₂Si+SiC_p)/Mg composites in the conditions of different loads and abrasive sizes has been presented. Their worn surfaces have been observed by SEM and extracted technology. The results show that the wear resistance of (Mg₂Si+SiC_p)/Mg composites is much higher than that of pure Mg, because the firm Mg₂Si and SiC_p reinforced phases could restrain the occurring of plastic deformation, enhancing the ability of wear resistance. The wear resistance of (Mg₂Si+SiC_p)/Mg composites increases with increasing the content of SiC particle. After hot-extrusion, (Mg₂Si+SiC_p)/Mg composites exhibits the better wear resistance. And the failure mechanism of (Mg₂Si+SiC_p)/ Mg composites mainly is abrasive wear.

The influence of MgCO₃ on the microstructure of AZ31 magnesium alloy has been investigated.The results show that the morphology of β-Mg₁₇Al₁₂ Phase in the alloy changed from continuous network to diseontinuous one, and their amount deereased with inereasing MgCO₃ content. AZ31 magnesium alloy has the best refining effect when the AZ31 magnesium alloy with addition of 0.6% MgCO₃ in 760 ℃, the average grain size of the α-Mg grain decreases from about 570 μm of AZ31 base alloy to 100μm, the reduced extent is about 82.5%, tensile strength are improved by 57.9%, the elongation percentage increases to 11.1%. Based on the analysis of EDS, Ebind and Gibbs free energy, the mechanism of grain refiner was generation of Al₄C₃ that can be used as crystallization center of the Mg when MgCO₃ added. Due to the generation of Al₄C₃ in AZ31 magnesium alloy, the grain boundary is pined and the grain growth is limited.The Al element was quickly pushed to the grain boundary along with the solid/quid interface, and generated the β-Mg₁₇Al₁₂ phase along the grain boundary which has an effect of immobilizing the grain boundaries. The distribution of alloying elements have changed.

To overcome the shortcomings of the traditional risk priority number(RPN) method, an evaluation method with multi-assignments was proposed based on the multi-expert evaluation. The expert weight was assigned by an abjective weight measure. The interval RPNs were sorted using the distance measure based on the idea of the technique for order preference by similarity to an ideal solution(TOPSIS). As an example, the RPNs of the fauer modes of 3 critical components of a certain type of machining center were calculated and sorted by the proposed method, and the weak links of reliability were found. The method can also be used to analyze the RPN of the complete machine.

The reliability analysis was performed for the related failures using the Copula theory, selecting the Gumbel Copula function in the Archimedes function group as the linking function of the CNC machine tool to establish the relationship between the complete machine and its subsystems. The parameters of Copula function were estimated by the genetic algorithm. Comparing with the series model, the failure analysis based on the Copula theory is more accurate.

Considering the random characteristics of the product disassembly process, introducing the defined concept of the cost disassemblability degree, a theoretical model was established for the assessment of the product disassembly cost. The probability distribution function of the disassembly cost was determined using the maximum entropy principle, and its solving algorithm was designed. A simulation was performed to verify the rationality of the proposed model and the effectiveness of the corresponding algorithm.

The load spectra of the wheel loader axle shaft were compiled by collecting the measured shaft load signals in the typical operation conditions to provide a reliable load condition for the prediction of the shaft fatigue life. Based on the compiled load spectra and the finite element analysis of the shaft, the fatigue life of the shaft was predicted using the Miner criterion. The effects of the operation conditions on the shaft fatigue life were analyzed. The results showed that the percentage of the fatigue damage is 68.6% in the operation in the cases of primary soil and semihumid soil, and it is less in other operation conditions. Considering the effect of the uncertainties on the fatigue life,the sensitivity analysis was performed on the shaft fatigue life. According to the sensitivity analysis result,the fatigue life increases by 3.7 times when the surface residual compressive stress reaches 100 MPa, and the fatigue life reduces by 80% when the load amplitude increases by 10%. The presented approach provides a reference for predicting the axle shaft fatigue life and improving its structural design.

Based on the D-type knotter, a mandatory strip type knotter was designed using the principle of reverse engineering. The performance of the knotter was simulated by the Pro/E software. The designed mandatory strip type knotter is mainly consisted of an oscillating roller cam mechanism, an incomplete olecranon gear mechanism, an incomplete primary rope folder bevel gear transmission mechanism, and a secondary rope folder helical gear and screw transmission mechanism. Based on the reversed data and according to the analyses of the kinematics and the gear tooth profile meshing characteristics, the axic inclination angle of the incomplete olecranon gear mechanism, the incomplete primary rope folder bevel gear transmission mechanism, and the secondary rope folder helical gear and crew transmission mechanism is 96°,96°,and 80°,respectively. The cam contour of the designed driven gear cam structure is formed by 2 circular arcs with radiuses of 28 and 18 mm and their tangent. The results of simulation by Pro/E showed that the designed mandatory strip type knotter knots steadily without interference, the structure of the driving gear cam and the alecranon gear cam surfaces changes smoothly, the cam follower can achieve the scheduled move and stop without jump, the main mechanisms operate at the appropriate time sequence, and in summary, the requirement of the knotter practical operation can be satisfied.

Using experimental optimal design method, nine concave surface morphologies were designed on ring model with different concave sizes and distributions. Laser graphic texturing was applied to treat process the bionic concave surface micro-morphology on the ring model. The anti-wear capacity of the ring model was investigated. Under the same conditions, comparative experiments were conducted on smooth model and concave model. The wear values and wear rates were determined by weighing the samples before and after the experiments. The wear values and wear rates of ten different samples were analyzed. By orthogonal experimental design, the range analysis was adopted to evaluate anti-wear properties the nine bionic samples. Under lubrication condition, the anti-wear properties of the concave models were better than that of the smooth model. The morphological parameters of the best anti-wear bionic model were obtained as follows; the concave diameter is 100 μm, the transverse space between two concaves is 450 μm, and the longitudinal space is 350 μm.

Three bionic models were proposed by modifying the NACA 0018 airfoil model according to the serrated trailing edge and nonsmooth leading edge of owl wing, named serrated trailing edge model, scalloped leading edge model, serrated trailing edge and scalloped edge combining model. The aerodynamic noise generated by the three models is investigated by numerical simulation. The source fluctuations of the flow field are computed by Large-Eddy Simulation (LES). Then the fluctuations are fed to the followed acoustical computation as input data. The far-field sound level is computed by FW-H equations. The results at small angle-attack, α = 6° and at chord Reynolds number Re = 1.6×105 indicate that all three bionic models outperform the NACA 0018 airfoil model in aerodynamic noise reduction. The serrated trailing edge and scalloped leading edge combining model performs the best among the three models in noise reduction.

A circumferential rigidity model and a circumferential inner force model of the shield segment were built based on the elasticity theory, and some related parameters of the segment joint were analyzed. The results showed that the inner forces calculated by the theoretical model coincide with the results by the finite element method, demonstrating a high accuracy. The rigidity model can reveal exactly the circumferential rigidity distribution in the shield segment structure. The maximal circumferential positive bending moment increases along with the reduction of the segment joint rigidity. The distribution of the circumferential bending moment changes with the changes of the joint number and the K block position.

Through the researching and comparison, mechanics parameters of keloids and normal skins were acquired by the relaxation experiments, creep experiments,and the tensile Tests.The results showed that there were significant differences between the elasticity modulus of the keloids and normal skins . Evaluation standards of the keloids mechanical properties were offered. It provided a reference for clinical researching and treatment of keloids.

A parallel algorithm of multiple ant colonies on Compute Unified Device Architecture (CUDA) computation model of CPU is presented. Several integrated different kind of ant colonies share common pheromone matrix during the process of solution construction and pheromone trail update steps. Multiple ants, corresponding to those iteration-so-far-tours, update pheromone matrix to obtain the diversity throughout explorative search phase; and the one, corresponding to the best-so-far-tours, deposits pheromone to accelerate constringency from exploitation phase. Ant colonies are mapped to the thread blocks and ants within colony correspond to massively threads of CUDA. The mix of MMAS and ACS is implemented that a new initializing method and the max and min pheromone trail limits are used. Convergences both in value and solution are proved and experiments on TSPLIB demonstrate that this algorithm can get good average solution quality, and outperform the sequential MMAS, ACS and parallel algorithm with a dual-core CPU.

A measuring method of software dependability based on Pi calculus was proposed. In this method, the running behaviors of software were decomposed into parallel or series of actions. Then the dependability of the software was quantitatively analyzed by comparison of the real running results with the expected results. Experimental analysis shows that the proposed method can simplify the measuring process of software dependability and timely and effectively monitor the changes of the software dependability. The results suggest that the proposed method can be applied for real-time measurement of software dependability.

An Adaptive Multi-population Genetic Algorithm (AMGA) was proposed, which was applied to feature selection and to find the optimal feature subset from high dimensional feature sets. AMGA consists of a Multi-population Planning (MPP) module and a Dynamic Selection Algorithm (DSA), both being self-designed. The proposed method extends the search space and adjusts the running states of multi-population, thereby controlling the premature convergence and increasing the local search capacity for the premature convergence. Segment from UCI and StatLog data sets were used to evaluate the proposed method. Results show that, comparing with standard genetic algorithm, the proposed method obtained an ideal result with less selected features and higher classification accuracy. The proposed method can be widely applied in the field of feature selection.

An approach based on Go knowledge to calculate the static evaluation value is proposed, and the parameter of the model is further optimized by genetic algorithm. Through this approach, various static evaluation models based on different levels can be obtained. The results of text show that, compared with the original model, the accuracy and operation speed of the proposed approach are promoted by 93% and 35% respectively. We expect that, with such operation capacity, this model can be applied in the module of the middle game and end game of computer Go. This model has a practical utilization in researches on computer games, artificial intelligence and game software.

Based on the characteristic quantification models of behavior, an unsupervised classification algorithm was used to analyze and describe the daily pattern of the elderly behavior. Based on the immune principle the classification method was optimized and improved. Results of contrast experiments show that the proposed method is effective to overcome the shortcomings of the traditional clustering method, such as slow convergence and extreme points, thus realizing a better classification rate.

A MiR-CR method was proposed based on conserved regions of influenza virus genes to discover human encoded miRNAs that regulate influenza virus. HA, NA, NP, PB2 and PA are taken as examples, seven human encoded miRNA are discovered. The result needs further experimental verification. Because the research of the miRNAs influenza gene expression in the infection is still in the early stage, the result of this work is of importance to discover the changing regularity of influenza pathogenicity and infection and apply in drug molecular design.

Based on the maximum inter-class margin of Support Vector Machine (SVM) and the minimum intra-class volume of Support Vector Data Description (SVDD), a discriminant algorithm is proposed, named Support Vector Data Description Discriminant Analysis (SVDDDA). This algorithm establishes two different concentric hyperspheres. The positive class samples are packed in the small hypersphere and the negative class samples are excluded from the large hepersphere. The objective function of the model maximizes the inter-class margin and minimizes the volume of the small hypersphere simultaneously. The projection coordinates are defined by the distance between the sample and the center of the hyperspheres. SVDDDA can preserve the inter-class discriminant information and intra-class scatter distribution. Results of experiment on public facial expression database demonstrate the efficiency of the proposed method.

In the scene with multiple sensing tasks in wireless sensor network, an Adaptive Capacity Control Strategy (ACCRP-MST) is proposed to use intermediate nodes efficiently. ACCRP-MST finds multiple node-disjoint paths for each source-sink pair by taking into account the tradeoff between residual energy and hop count and achieves traffic load balance. Then it uses the Multi-path Source Routing (MSR) protocol to report the acquired data in order to save the storage and computing cost of the intermediate nodes. Simulation results based on NS-2 show that the proposed ACCRP-MST outperforms several other energy-efficient routing protocols by increasing the network lifespan and communication reliability and efficiently enhancing the overall performance of the network.

An Energy-Efficient Data Gathering Algorithm (EEDGA) is proposed for monitoring events in a kind of reactive Wireless Sensor Networks (WSNs). First, EEDGA selects a clusterhead according to the correlation and residual energy among the nodes located in the event area. Then, EEDGA utilizes ant colony algorithm to form a chain of data gathering inside the cluster. EEDGA also proposes a simple method to reduce the energy consumption by choosing active nodes according to the demand of measuring accuracy. Simulation results show that the proposed EEDGA can achieve higher energy efficiency and longer network lifetime than that of the traditional routing algorithm.

To deal with the disturbances of a kind of uncertain nonlinear systems, a dynamic disturbance observer system is constructed. According to this disturbance observer system, a fast neural network disturbance observer is proposed. Based on the observed disturbance error the observer approaches the unknown synthetic disturbance by on-line adjusting its weights. An adaptive terminal sliding mode control is designed based on neural network disturbance. It is strictly proved that the states of a closed system converge to zero within finite time. Therefore the convergence speeds are improved under this scheme. Simulation results demonstrated the effectiveness of the proposed scheme the capability of the observer to approach disturbance.

A control structure is proposed for trajectory tracking control of nonholonomic mobile robots. It integrates the backstepping kinematic controller and a sliding mode controller with Adaptive Dynamic Recurrent Fuzzy Neural Network (ADRFNN). The genetic algorithm is used to optimize the parameters of kinematic controller that effectively suppresses the excessive initial speed and output torque caused by large initial error of posture. The ADRFNN is developed to achieve on-line estimation of the part of dynamic nonlinear uncertain, which greatly reduces estimation errors of uncertainties. By combing ADRFNN with the adaptive robust controller, this method can not only solve the problem of parameters and non-parameter uncertainties of mobile robots, but also eliminate input chattering of the sliding mode control. The stability and convergence of the control system are proved by Lyapunov theory. Simulation results demonstrate the effectiveness of the proposed method.

A high-precision nanosecond pulse driver was designed and developed by using DSP and analog proportional integral (PI) control technology. In terms of hardware, this system includes nanosecond pulse generator circuit, constant current control circuit, slow starts circuit, ESD protection circuit, over-current protection circuit, et al. Meanwhile, in respect of control scheme, the current stability was effectively ensured by use of simulation PI algorithm for feedback control. The driver with pulse width of 15 ns, peak drive current of 20 amps and pulse rise/fall time of less than 5 ns is achieved. With this driver, performed driving test on a QCL with a center wavelength at 7.71 μm, which is manufactured by institute of semiconductor, CAS. The results show that during long term (>100 h) operation, the center wavelength of the tested QCL does not shift, and there is no glitch regarding the driving current. Therefore, the fabricated driver provides superior performance that guarantees the application of the QCL in infrared gas detection.

A model was built for the calculation of temperature of the insulated grid bipolar transistor(IGBT) using the lumped parameters based on analyzing the heat transfer mechanism in the IGBT. The solution space of the inverse heat transfer problem was discreted by the boundary element method. The inverse heat transfer problem was solved by the conjugated gradient method to get the accurate equivalent thermal resistance and capacitance. The results by the proposed method were compared with the experimental data of the junction temperature of IGBT provided by the manufacturer technical documents as well as with the results by the finite element method, the temperature calculation error is less than 5%.

Aiming at the combination explosion problem induced by choosing root nodes in the extraction of functional regularity in digital ICs, an algorithm capable of reducing the number of root nodes by extracting the chain-like frequent subcircuits is proposed. By establishing sequence-dependent edge-weight model, the small chain-like frequent subcircuits can be extracted fast. Furthermore, to improve the efficiency of regularity extraction, the nonfrequent root nodes can be deleted gradually by utilizing structure dependencies between small chain-like frequent subcircuits and other structure templates at gate level, which avoids the repetitive extraction of small frequent subcircuits. Experimental results show that the proposed algorithm can solve the combination explosion problem effectively, extract the high frequency candidate subcircuits with high priority and reduce runtime of regularity extraction observably.

The theory of multidimensional vector orthogonal transformation matrix was extended. A new 4-D vector matrix Discrete Cosine Transform (DCT) operator was presented based on the theory. In addition, the orthogonality of the operator and energy concentration were verified. Also a technique was proposed to generate the quantization cube suitable to Multidimensional Vector Matrix DCT (MD-VMDCT) transform codec. A sampling model for 3D Vector Matrix DCT (3D-VMDCT) codec was also proposed, which can eliminate spatial redundancy and color redundancy among the pixels in the color image by use of the supper concentration of the MD-VMDCT. Then quantization and Huffman coding were carried out to eliminate statistical redundancy. Experimental results show that the algorithm is correct and effective, and under the same conditions it outperforms the Joint Photographic Experts Group (JPEG) method.

In this paper the statistic distribution property of Synthetic Aperture Radar (SAR) image was analyzed. Then a fast Otsu segmentation algorithm for SAR images was proposed, which satisfies the requirements of SAR automatic target recognition on SAR image segmentation. In this proposed algorithm, the Constant False Alarm Rate (CFAR) technique is first employed for coarse segmentation, and then Otsu was adopted for fine segmentation. Experimental results show that the proposed algorithm is not only efficient and accurate, but also can also achieve higher objective evaluation values.

A new algorithm based on the 1-Dimensional Empirical Mode Decomposition (1-DEMD) is developed to extract details from 2-dimensional images. First, the 2-dimensional image matrices must be previously transformed into 1-D image vectors according to the line or column of the image matrix; then the 1-DEMD can be used for decomposition. In order to achieve both vertical and horizontal details of the image, matrix-to-vector transform according to the line or column of the image matrix is cross-used by 1-DEMD. Experimental results validate the feasibility of the proposed method.

Based on the research of image structural properties in DCT domain, DCT Domain-based Structural Similarity (DCTSSIM) was proposed. The method takes DCT domain information as the structural information and weights the DCT coefficients by different sensitivities of human visual system. Experimental results illustrate that, compared with SSIM and Peak Signal to Noise Ratio (PSNR), the proposed method is more consistent with the subjective assessment of human beings for JPEG2000 and JPEG images, thus can be used to describe the visual perception of image effectively.

A novel Suboptimal Fading Square Root Central Difference Kalman Filter (SFSRCDKF) algorithm is presented. This algorithm can enhance the observability, increase the convergence speed and improve the locating accuracy in the Single Observer Passive Location (SOPL). It uses adaptive suboptimal fading factor restricted by the orthogonality principle to real-time adjust the gain matrix, which can ensure the orthogonality of the new observation residuals at different time. Therefore, the response speed and the adaptive correction capability of the filter are improved when the state changes or the estimation is biased. The new filter also uses square root of the error covariance instead of the error covariance involved in filtering to ensure numerical stability. Simulation results show that, under different conditions, the new algorithm performs more stably with higher convergence speed and higher locating accuracy.

Communication networks which carry M2M and H2H services are modeled as a session level, discrete time tandem queuing network to explore how the vast M2M small data services can affect the H2H services. In probability generation function domain, we propose a method to map the service process of M2M traffic into the service time of H2H traffic and get the equivalent expression of actual service procedure, which simplifies the solving procedure of tandem queuing network with mixed traffic arrival and heterogeneous servers. By solving the tandem queuing system, the relationship between the performance metrics of H2H services and arrival rate, service rate of M2M service is revealed and depicted. Admission control methods used to M2M small data service are investigated.

The optimization model based on assignment problem is implemented to the Peak-to-Average Power Ratio (PAPR) mitigation of Cognitive Radio (CR) systems, which is based on WOFDM. Under the condition that the number of the wavelet filter bank levels is fixed, the optimal wavelet packet basis is chosen to obtain the optimal performance of the system PAPR. Analysis and simulations show that implementation of the optimization model can not only ensure the Bit Error Rate (BER) performance of the system, but also mitigate the PAPR of the system without requiring any side information. In addition, the design of the new algorithm satisfies the constraint of parallel computation, so parallel computing can be implemented based on Multi-core Programming (MP). The execution performance of the new algorithm can be ensured.

Based on the time domain analysis of the continual signals which are produced by atan phase detector in the carrier phase-locked loop(PLL), Some physical laws about these signals are given. Then study the influence of the pre-detection integral time on the bandwidth of phase detector in carrier PLL, and derive a formula expressing the relationship between theoretical maximum value of the bandwidth and the pre-detection integral time. Finally, the theoretical simulation and experiment results confirm that the research conclusionis accurate.

Egg powder with high content of Zn was prepared by vacuum freeze-drying technique. Zinc citrate was taken as additive for Zn enrichment of the egg powder. The solubility of powder was investigated. On the basis of single experiment, the effects of the parameters of drying pressure, powder thickness, freeze-drying time on the solubility of the Zn enrichment egg powder were investigated with second general combination of rotating design method. The regression model of the solubility of the Zn enrichment egg powder as function of these parameters was obtained. Results show that influences of powder thickness and freeze-drying time on the solubility are significant (p<0.01). The impacts of these parameters are in the order: powder thickness, freeze-drying time and drying pressure. The optimum parameters are powder thickness of 4.3 mm, freeze-drying time of 12.51 h, drying pressure of 43.6 Pa. Under such optimal condition, the wetting sedimentation time is 171.27 s, disperse time is 86.53 s, the solubility is 76.97 g/100 g, and comprehensive score is 109.23. To compare the nutritional compositions with fresh egg, the Zn gain of the enrichment egg powder can reach 73.05%.