For four-wheel independently drive in-wheel motors electric vehicle, the dynamic control systems, such as direct yaw moment control, can be easily achieved. Accurate estimation of vehicle state variables and uncertain parameters can improve the robustness of the vehicle dynamic control system. Various sensors are generally equipped to acquire the vehicle dynamics. For both technical and economic reasons, some fundamental vehicle parameters, such as the sideslip angle and tire-road forces can hardly be obtained directly by sensors. Therefore, a state observer is developed to estimate these parameters based on Unscented Kalman Filter (UKF). In addition, a nonlinear dynamics tire model is utilized to improve the accuracy of tire lateral force estimation. Compared to the reference model built in AMEsim, the simulation results show that the proposed observer can provide the precision values of the vehicle state.

To meet the vehicle performance constraints, a design and analysis model of the powertrain of Plug-in Hybrid Electric Vehicle (PHEV) was proposed. In this model the manufacturing cost of the vehicle dynamic components and the annual usage cost were taken as the evaluation indices, and the car users' habits were classified. According to different design requirements of All-electric Range (AER), the powertrain performance parameters of PHEV were optimized. The results show that, compared with the original vehicle, the total annual optimization cost of the designed PHEV for different AER is reduced by 6.0% ~ 20.0%. The best optimization design requirement for AER is 40 km. Compared with other AER and the original vehicle, the annual optimization cost can be reduced by up to 24.9%.

To solve the problem of high emission in coldstart process of auto-transmission vehicle, the vehicle coldstart condition, idle braking and starting acceleration conditions were investigated. Then, a engine warm up recognition strategy to judge the vehicle coldstart, and a neural control strategy to reduce idle emission were proposed. Also the strategy to keep first gear in coldstart to increase the air intake and fuel injection was proposed to accelerate the catalytic ignition. This method was tested in emission laboratory. Results show that the catalyst can accelerate the light-off, and the emission in coldstart process is significantly reduced.

In order to accurately detect the faults of motor response of hybrid vehicle and determine the fault degree in time, the fuzzy logic arithmetic was introduced to the fault detection. The method to obtain the threshold of motor load difference value was analyzed on the promise of considering the influences of the motor speed and the load. The motor working status was illustrated based on the threshold. According to the characteristics of the input and output signals, the membership function of the variables was established. The evaluation principle based on fuzzy logic was proposed, and a motor fault diagnosing model was built in Matlab/Simulink. Simulation results verify the accuracy and practicality of the model.

An obstacle detection method based on 3D laser scanner and range image is proposed for intelligent vehicle. First, the range image of one scan is established based on the Welodyne HDL-32E laser scanner data. Then, according to the range image analysis, the point cloud in one row of the image is classified into several groups. Finally, the point clouds in different groups are detected by a linear model in column. Simulation reveals that using the proposed method, the rate of false detection caused by the obstacle lock is reduced, and this method well adapts to the terrain change.

The dynamic model of eight Degrees of Freedom (8D) Four-wheel Steering (4WS) vehicle was established. Using the theory of artificial neural network, a Radial Basis Function Neural Network (RBFNN) controller was designed and it was trained by the direct close-loop training method off-line. In order to overcome the shortcomings of poor adaptability caused by off-line training, a single neuron adaptive PSD controller was also designed. The RBFNN controller and the single neuron adaptive PSD controller composed a compound control system to jointly complete the steering control of 4WS vehicle. Simulation experiments were carried out by combining the RBFNN compound controller with the 8D 4WS vehicle dynamic model. Results show that the designed RBFNN compound controller can effectively improve the steering stability of 4WS vehicle both in constant speed and variable speed conditions.

The optimal control of Hybrid Electric Vehicles (HEV) is a problem of multi-objective optimization. An optimization model was built based on the working characteristics of the engine, the Motor/Generator (MG) and the battery pack, and their interactions. The optimal control problem was divided into three levels, the energy management, power distribution, and coordinated control. The optimal results were applied in the real-time control process using the hierarchic control method. The dynamic simulation model was built and the on-line test was carried out on dSPACE platform. The results verify the feasibility and effectiveness of the hierarchic control method. Compared with the rule-based method, the proposed method can improve the drivability by 22.2%, the fuel consumption by 10.6% and the electricity supply performance by 15.5%, respectively.

A construction procedure of driving cycle is proposed using K-means clustering method. Experimental data of a city driving cycle is obtained by driving test vehicle, and is divided into "micro-trip" after preprocessing. The micro-trips are then clustered into groups in "average speed","distance" and "cruise time percentage" using the K-means clustering method. The micro-trips are pick out on the principle of the closer distance between the micro-trip's driving feature and the center of cluster, which is more representative of the cluster. Then the final city driving cycle is developed. The data of fuel consumption and pollutant emission are collected using revolving drum test and exhaust-gas analysis. The characteristic parameter, fuel consumption and pollutant emission among different typical driving cycles are compared. Results show that the driving cycle constructed by this new approach can better reflect the city actual traffic conditions and it has high practical value.

To analyze the applicability of four vehicle delay models for signalized intersections, the experimental environments were designed on the basis of certain traffic demands and signal control. First, the degree of difficulty for using the four vehicle delay models was compared in view of model structure and the number of variables. Then, for the microscopic traffic simulation software VISSIM, the impacts of the travel time section and analysis duration on the measured delay were discussed. Finally, several experimental scenarios were designed according to the cross-classification of the intersection degree of saturation and cycle length. The goodness of fit between each vehicle delay model and the simulation model was investigated. Results show that the Webster model is easy to use and apt to underestimate vehicle delay and its applicability is the worst; the ARRB model is more complex and its applicability is the best; the HCM1985 model is also easy to use and it applicability is moderate; the HCM2000 model has broad application range but it is too complex, and its applicability is moderate. On the whole, for isolated signalized intersections, the ARRB model displays the best performance in vehicle delay estimation. This research may provide a theoretical basis and reference to the selection of vehicle delay model.

A forecasting model was proposed based on the short-term changing characteristics of Available Parking Space (APS). This model integrates the wavelet analysis, Particle Swarm Optimization (PSO) and Wavelet Neural Network (WNN). First, the APS time series were decomposed and reconstituted by wavelet analysis. Then, WNN model was used to forecast the reconstructed time series respectively. The PSO method was employed to optimize the selection of the initial parameters of the neural network. Finally, the final forecasted ASP was induced by integrating the prediction results. A case study was carried out to verify the applicability of the proposed model. Compared with simple WNN model, the new method enjoys higher accuracy and stable performance that the APS forecasting can be improved by 5 to 7 times by this new method.

The commuter's daily activity-travel schedule is proposed and its key time allocation is investigated. According to the comparison of hazard model and Support Vector Regression (SVR) model, travel time prediction models and activity duration prediction models are developed by employing SVR. Then, the continuous time allocation, i.e. all the travel times and activity durations in daily activity-travel schedule are derived and the transit priority policy is evaluated using these models. The results indicate that the model system has a high level of prediction accuracy, and the goodness-of-fit of SVR models is higher than that of Hazard models. This study provides useful insights into commuter's activity-travel time allocation decision. It also serves a foundation that future models of full-scale daily activity-travel pattern can be built on. Moreover, it provides potential for transportation demand management policy analysis.

In order to reflect the routing selection behavior considering reliability in essence, a mathematic model of optimal-reliable routing in stochastic and dynamic traffic network is developed based on reliability theory. First, through defining reliability as the objective function of routing, the adaptive reliable shortest path problem in stochastic and dynamic network is established to reflect the stochastic, time-varying and risk characteristics of travel time. Then, using optimization theory, the adaptive reliable shortest path problem is transformed the dynamic programming problem. Finally, an algorithm based on dynamic programming is developed to solve the proposed problem. Numerical results in typical transportation network demonstrate the validity and feasibility of the proposed algorithm.

In order to find traffic conflict discrimination indicator, which can reflect the driver's mental load, and to establish the traffic conflict discrimination method based on the indicator, the correlation between the conflict stimuli and the pupil diameter is determined by comparative analysis. A pupil diameter based construction conflict self-feedback discrimination method is proposed based on the research of the influence of traffic conflict on the pupil diameter. This method overcomes the shortcomings of normal template discrimination method, including too much manual intervention and lower discriminating efficiency. Construction conflict discrimination results show that the error rate of the self-feedback method is 5.56%, and the error rate of the non-feedback method is 24.4%, thus, the high discrimination precision and no manual intervention of the self-feedback discrimination method are testified. The proposed method can meet the present requirements of traffic conflict discrimination and road safety evaluation. This work lays foundation for further study of relationship between traffic conflict severity and pupil diameter, and to build traffic conflict quantification system based on driver's eye movement characteristics.

The effect of filler-to-bitumen ratio on the low-temperature cohesive strength of SBS modified asphalt mortar was studied using quantitative testing technology. Two control variables, filler-to-bitumen ratio and temperature, were selected in the experiments. Results show that the change rule of the low-temperature cohesive strength is significantly influenced by both the filler- to-bitumen ratio and the low temperature environment. Appropriate filler-to-bitumen ratio can improve the low-temperature cohesive strength of asphalt mortar. Excessive content of mineral power fillers will produce filler agglomeration, causing additional damage points, which lead to significant decrease in the strength. The cohesive strength of the asphalt matrix increases as the environment temperature decreases. However, when the temperature is below the asphalt's embrittlement point, the asphalt matrix can be embrittled, which significantly attenuates the cohesive strength of the asphalt matrix. Therefore, the cohesive strength of the asphalt matrix, its embrittlement state and the effect of the contact interface strength of asphalt filler should be comprehensively considered in determining the optimal filler-to-bitumen ratio. Moreover, these factors should be combined with local low temperature conditions.

In order to analyze the static response of the carbon fiber-steel-concrete composite beams under vertical concentrated load, an analytical solution of the double joint layer composite beams under vertical concentrated load was derived. This analytical solution is based on classical theory of elasticity. The equation of deflection curve of the carbon fiber-steel-concrete double joint layer composite beams was obtained. Numerical calculations of the composite beams with or without interaction are carried out to compare the effect of interaction. The results demonstrate that the analytical solution is correct and reasonable, which provides the engineering designers with a reliable calculation method and has strong practical value.

In this paper, the influence of mineral admixtures on the rheological properties of fresh Self-Compacting Concrete (SCC) mixture was studied using a new type of concrete rheometer. The results show that both yield shear stress and plastic viscosity of the SCC mixture decrease as the content of fly ash in the mixture increases. As the content of Ground Granulated Blast Furnace Slag (GGBFS) increases, the thixotropy behavior of the SCC mixture becomes more obvious, and the yield stress decreases first and then increases. Both the yield stress and the thixotropy behavior of SCC mixture increase with the addition of limestone powder. For all the SCC mixtures with different mineral admixtures, as the storage time increases, the thixotropy behavior of fresh concrete is gradually weakened, the yield stress increases and the plastic viscosity decreases.

A three-dimensional Finite Element Model (FEM) of centrifuge was established first, and the reliability of the model was validated by centrifuge test. Then this method was compared with the current specified m-method. A case study of cantilever row piles was carried out to analyze the impact factors to the internal force and head horizontal displacement of the pile. These impact factors include the cross-section area of the pile, the pile spacing and soil strength. Results show that the design of row pile retaining structure by the m-method is conservative. The three-dimensional FEM method is more consistent with the centrifuge testing results than the m-method. The bending moment and the horizontal displacement of the pile linearly increase as the pile spacing increases and the pile width decreases. They nearly exponentially increase with the cohesion and internal friction angle of the soil.

A numerical simulation model for vehicular cooling modules was established and verified by experiment. On this basis, the numerical study of the vehicular detached cooling modules was carried out. First, the detached cooling modules with serial structure were set up, the influence of the position and number of air intakes on the cooling performance was investigated. Then, the detached cooling modules with non-serial structure were established, the influence of the relative position between Heat Exchangers (HEs) and fan was taken into account. The results show that, when the cooling modules were serially connected, the increase in air intakes has no obvious impact on the cooling performance. When the cooling modules were non-serially connected, the cooling performance is greatly improved. In addition, the cooling efficiency of the scheme with double HEs is higher than that with single HE.

Based on steady-state heat metering method, a measurement system is developed, which can be used to test the effective conductivity of a granular system under different temperature and vacuum degree. Using this measurement system, experiments for HIT-LS1# lunar soil simulant are conducted. The results show that the effective thermal conductivity of the lunar soil simulant increases with temperature, and decreases with the increase in vacuum degree. Vacuum degree has on obvious impact on the effective thermal conductivity, especially in the case of high vacuum.

A cartridge one-way relief valve is proposed, which has the characteristics of direct detection. The cartridge valve has the features of small override pressure and high precision of control pressure. To analyze the mechanism of the regulating property, a transfer function is established according to the theoretical analysis of the relief process and the corresponding key parameters are qualitatively analyzed. The HCD simulation model of the relief valve is established using AMESim software. The effects of the key parameters on the dynamic characteristics and the steady-state performance of the cartridge one-way relief valve are quantitatively analyzed. Theoretical analysis and simulation results indicate that the direct detection characteristic forms the interstage negative feedback, which can effectively reduce the steady-state override pressure. Furthermore, adjusting the size of the annular gap helps to improve the dynamic characteristics of the relief valve, and adjusting the mandrel end surface area helps to improve the steady-state characteristics of the relief valve. The one-way valve with this structure helps oil supply.

To improve the transmission quality and bearing capacity of orthogonal variable transmission ratio face gear, the influences of the geometric parameters of gear pair on the contact ratio are investigated. According to the spatial gear meshing theory, the meshing coordinate is established and the equations of the section curve and crest curve s are deduced. The orthogonal variable transmission ratio face gear is transformed to non-circular gear rack and pinion. The calculation method of the contact ratio is proposed, and the influences of the geometric parameters of cutter gear and gear pair on the contact ratio are analyzed. The proposed calculation method of the contact ratio is verified by the measurement experiment of the actual contact ratio.

In order to save energy in hydraulic excavator, the excavator working condition is analyzed and an electro-hydraulic energy regeneration system is designed. To theoretically analyze the characteristics of the energy regeneration system, a simulation model is built and numerical analysis is carried out. To verify the simulation results, the energy regeneration system is equipped in a 20-ton excavator and large number of tests is conducted. Both simulation and test results show that in single operation, the maximum energy regeneration efficiency of the electro- hydraulic energy regeneration system can reach 40%. When equipped with this system, the excavator can save energy by 17.6% without loss of its original operating characteristics.

To improve the accuracy preservation of three-jaw chuck, the effect of the abrasion of chuck components on its centering accuracy was investigated. The model of centering accuracy and repeatability of three-jaw chuck was developed. Results show that among the position errors of jaws induced by abrasion of chuck components, the radial position error dominates. For the wedge type power chuck, the repeatability accuracy is equal to the interval between chuck body and wedge plunger. Only when the interval between chuck body and wedge plunger is limited within the double systematic error of the centering accuracy, the method of turning top jaw can renovate the centering accuracy after long-term use of the chuck. But the accuracy inefficacy caused by the enlargement of the interval between chuck body and wedge plunger due to the abrasion cannot be renovated by turning the top jaw.

In order to study the secondary deep drawing forming theory, the tensile stress theoretical model of secondary deep drawing is proposed. First, by theoretical analysis, the calculation formula of tensile stress is deduced, and the calculation method of the limited secondary deep drawing coefficient is established based on the Swift instability theory. Second, the accuracy of the calculation model is examined by secondary deep drawing simulation, and the applicability of the calculation model is verified by means of successful secondary deep drawing forming experiment. The results show that the theoretical model of the tensile stress and the limited secondary deep drawing coefficient has high precision.

The influences of tool temperature and stress state of stamped part on its hardness distribution are investigated in tailed die quenching process of vehicle-use ultra high strength steel BR22MnB5. The results show that the hardness of the stamped part decreases with the increase in tool temperature. When the tool is heated up to 400 °C, a part with hardness of 300 HV in the hot region and 460 HV in the cold region is produced. Furthermore, during the stamping, the inner fillet of the stamped part is compressed while the outer fillet of the blank experiences tensile stress. So the hardness of the outer fillet is higher than that of the inner fillet in the stamped part.

The quantitative relationship between average grain size and simple pass rolling deformation of magnesium alloy is studied. The concept of shape factor is introduced and the regression equation of them is calculated. The effect of β phase on the deformation property of AZ31 alloy is mathematically analyzed. The β phase can be deleted by homogenization treatment to obtain the accurate regression equation between grain size of AZ31 alloy and simple pass rolling deformation. This study may provide a direct reference to simplify the rolling production.

The ceramic coating fabricated by micro-arc oxidation at different current density is employed to improve the corrosion resistance of Friction Stir Welding (FSW) welds on aluminum alloy. The morphology and phase composition of the coating were investigated by Scanning Electron Microscopy (SEM) and X-ray Diffraction (XRD). The surface roughness of the coating was characterized by Confocal Laser Scanning Microscopy (CLSM). The corrosion properties of the FSW welds before and after micro-arc oxidation were analyzed by Electrochemical Impedance Spectroscopy (EIS) and polarization curve (LSV). Results show that the thickness of the ceramic coating layer increases with the current density and it becomes more difficult to break down the coating. Once the coating is broken down a lot of melts will form and the surface roughness increases. Furthermore, the ceramic coating produced at current density of 10 A/m² shows better wear resistance than the others.

Laser welding-brazing of 5A06 aluminum alloy and galvanized steel sheet with filler powder was carried out. The obtained butt joint exhibited dual characteristics. The interfacial properties and microstructure of the joints examined using Scanning Electron Microscopy (SEM), Energy Dispersive Spectrometer (EDS) and X-ray Diffraction (XRD), and the tensile strength was measured. Results show that a non-uniform intermetallic compound (IMC) with serrated shape is formed at the interfacial layer, which is composed of Fe₂Al₅ and FeAl₃. The thickness of the IMC layer varies from 2.5 ~ 7 μm, and it increases with the laser power. At the top surface of the galvanized steel the IMC layer reaches the thickest. Using this laser welding-brazing method with special filler powder, the IMC thickness can be controlled to less than 10 μm, and butt joints with certain tensile strength can be obtained under appropriate welding parameters.

The performance of the relational classifiers based on homophily is poor for the classification in networked data with heterophily. To solve this problem, a relational neighbor algorithm based on propagation distributions is proposed for heterophilous networks. The algorithm adopts the second-order Markov assumption and considers the influence from the neighbors of the unlabeled nodes' neighbors. At the same time the algorithm propagates the influence between nodes through computing propagating class vector and propagating reference vector; and by means of combining the relaxation labeling collective inference method, the algorithm continuously updates the results until the class distributions converge. Experiment results show that the proposed algorithm performs better on the networks with heterophily.

The Neighborhood Valley-emphasis method can not get the right threshold value in some cases, such as the valley feature between the target and background is not very distinct. In order to solve this problem, a global thresholding method is proposed. This method is based on the gray information around the valley-point neighborhood and the relative characteristics between the valley point and its adjacent crest-point. The proposed method weights the objective function with the gray information around the valley-point neighborhood and the relation between the valley-point and its adjacent crest-point. It improves the accuracy of the threshold obtained by OTSU. The optimal threshold got by the proposed method has less valley-to-crest ratio. In other word, the valley gray is taken as the optimal threshold, which has larger height difference with it adjacent crest-point. In order to improve the efficiency, a recursive single threshold method based on the aforesaid algorithm is used to achieve the image multi-threshold segmentation. Experiment results show that the proposed method has great segmentation performance and low time complexity.

A Random Sliding Quantum Key Window (RSQKW) management algorithm is proposed in order to handle the problem of applying Quantum Key Distribution (QKD) in classical network encryption. First, a sliding window mechanism is introduced to control quantum key consumption. Then, the algorithm negotiates the control parameters of the sliding window by detecting the status of the quantum communication network and the final quantum key generated rate. Finally, using the stochastic characteristics of the quantum key to calculate the sliding step width, the quantum key series in the sliding window can be used as a new secret key for classical encryption algorithm. Comparing with traditional method, the RSQKW algorithm can better dynamically manage the quantum key and enhance the efficiency and effectiveness of quantum key distribution and management. Experimental results in a real quantum key distribution network demonstrate the feasibility and effectiveness of the proposed algorithm.

Objective retrieval plays an important role in daily life, however the model "Bag of words" exists some problems for image retrieval. For example, the quantified visual words reduce the discriminative power of the local features and the model does not capture the spatial relationship among local features, thus affecting the retrieval performance. In order to realize objective retrieval in complex environment, a feature saliency constraint based image retrieval method is proposed. By locating the salient object, this method can enhance the discriminative power of the local features and capture the spatial relationship among local features. First, the salient regions are distilled and the features inside these regions are marked as salient features. Second, the image retrieval is performed with spatial pyramid model using the salient features. Finally, the distance ratio of any two pairs of the features in each image is computed using the pair of salient features of the query image and retrieval image; then, the ratios which meet the threshold are recorded and their average is calculated to obtain the final results. Experimental results show that the proposed method improves the accuracy of the retrieval significantly and reduces the computing time during the geometric verification.

Most traditional face recognition systems can not be implemented for fast and accurate face recognition under cluttered background. To solve this problem, an improved Scale Invariant Feature Transform (SIFT) method and a fuzzy control strategy are proposed. First, four new angles are computed from the sub-region orientation histogram, which represent the orientation information of each SIFT feature. Then, the progress of face recognition is limited in a range based on the new angles, meanwhile the SIFT features are split into two types according to the size; only the features of the same type are computed, leading to significant simplification of the algorithm, thus a fast SIFT algorithm is obtained. Finally, a fuzzy closed loop control system is applied to increase the accuracy of face recognition, which leads to a decrement of the incorrect matching. The results show that the computing speed of the improved SIFT method is raised more than 40% comparing with the original SIFT algorithm and the recognition rate is raised 10% even under the clutter conditions where the illumination, posture or expression are changing.

To solve the problem of situation evaluation for network information content security incidents, a hierarchical situation evaluation model parameter calculation method are proposed, through the analysis on the multi-dimension characteristics of network information content security incidents. The model uses hierarchical structure and calculates the situation assessment values of incident level, area level and system level respectively. Among these levels, the incident level is based on the behavioral characteristics and content characteristics of the incident characteristics; the area level is according to the relationship characteristics and location characteristics, and system level integrates the involved area levels. The method to calculate the parameters is defined. Simulation results show that the model and method is feasible and reliable. It can reflect the impaction of incidents effectively and grasp the change rules of the incidents in situation evaluation of information content security incidents.

A novel differential evolutionary algorithm based on could population (CPDE) is proposed to solve the network security situation prediction. The proposed concepts of cloud population and the distribution chain promote the diversity of the population. In this algorithm, first, the intrusion operator is employed to introduce the competition among the cloud populations, where the winners will implant their distribution into other cloud individuals. Then, cooperative operator is used to introduce the collaboration among the cloud individuals and perform the differential operation. Finally, the accuracy of the algorithm is improved using the local search operator. Experiment results show that the proposed CPDE is an efficient high-dimensional evolutionary algorithm and possesses certain advantages in optimizing the prediction model of the network security.

To meet the real-time and accuracy requirements of collision detection in Human-computer Interface (HCI) system, a fast collision detection algorithm is developed based on cloud computing model. First, a new construction method of balanced OBB bounding box tree by split plane is proposed. Second, the concept of marked traversing tree is introduced, and it is marked in depth or breadth traversal using stack on OBB to reduce the number of intersections in collision detection. Third, the task tree is divided using Map-reduce Model (MRM), and it is executed in parallel for the sub-tasks after division by the cloud model, thus to reduce the detection time. Finally, each sub-task result is identified and logical operation on the identified sub-task is carried out; the operation result is used to determine whether collision has taken place or not. Comparative experimental results show that, compared with the classic I-COLLIDE, MPI and Pipelining algorithms, the proposed algorithm has obvious advantages in terms of efficiency and accuracy that can meet the requirements of real-time and accuracy of HCI in complex virtual space.

To address the sampling problem of protein conformational space, an Ab-initio dynamic-step-size searching method of protein conformational space is proposed. This method is based on the energy tree-based searching framework. The high-dimensional optimization space of dihedral angle is projected to a low-dimensional space of feature vector with the protein conformation feature extraction, effectively avoiding the curse of dimensionality problem. The feature space is discretized according to the energy and temperature. Then the layers are systematically divided into cells to reduce the searching space. Relevant Fragment Assembly (FA) step-size and Monte Carlo disturbance step-size are set according to the specific energy layer, and the corresponding Metropolis criterion is employed to accept the conformation within different temperature layers. The replica-exchange method is used as auxiliary method to enhance the sampling of native-like protein conformation. Test results of 12 proteins show that their native-like protein conformations can be reached successfully and effectively by the proposed method.

To resolve the impending deadlock problems in the path planning of Overhead Hoist Transporters (OHTs) in unified Automatic Material Handling Systems (AMHSs) of 300 mm wafer fabrication, an impending deadlock-free path planning strategy was proposed on the basis of the graphic theory. First, the impending deadlock problems were formally described. After introducing novel concepts of chains and share points, necessary and sufficient conditions were identified and an available approach was put forward using the graphic model. Then, a mathematic model was set up with an optimal objective function of minimizing the transporting time. Finally, by executing deadlock detection and resolution steps successively, an impending deadlock-free scheduling solution was assigned. Simulation experiments indicate that the proposed strategy can effectively detect and resolve the impending deadlocks occurring in OHTs' path planning during material handling processes. Thus, it can improve the security and stability of wafer fabrication systems.

In order to improve the resolution of low precision small photoelectric encoder without increasing the encoder size and the scribed line number, an interpolation chip is designed that is suitable to small photoelectric encoder. First, the advantages and disadvantages of current electronic interpolation methods are analyzed, considering the resolution, precision, the circuit complexity and integration, and other factors. Based on the theory of phase modulation, an interpolation algorithm is proposed, which converts the measurement of spatial phase displacement to that of time difference value of instantaneous cycle. Then, the overall architecture of the whole chip is designed according to the algorithm principle, and each module of the chip is designed using Cadence software. The instantaneous cycle value of the modulation signal is obtained by simulation of the overall chip circuit. Finally, the measured spatial angular displacement after subdivision is compared with the theoretical value, and the subdivision precision is calculated. Experiment results show that, when the frequency of the photoelectric encoder signal input is in the range from 1 kHz to 100 kHz, the chip can realize optical signal segment of 0 to 100 times. When the input is at 100 kHz, the precision of 0.4571' can be achieved. Compared with other similar processing circuit, the designed chip has the advantages of high integration level, combining the distinguishing function and subdivision, good portability, and has certain engineering application value.

To meet the accuracy requirement of surface three-dimensional (3D) Magnetic Resonance Spectroscopy (MRS) forward modeling. Hammer integration algorithm was proposed based on unstructured non-uniform tetrahedral mesh. First, finite element method was used to calculate the 3D magnetic field in the subsurface, and the sensitivity kernel expression was deduced. Then, to overcome the shortcoming of conventional mesh grid, that the precision becomes poor as grid size is large, while smaller grid size leads to enormous calculation task, Hammer integration was applied to calculate the kernel function with conventional grid. Synthetic results including 3D kernel function and aquifer models show that, using Hammer integration, higher precision can be achieved with fewer grids and nodes. The forward model can be calculated efficiently.

A detection algorithm of the presence of deception jamming is proposed based on information dimension with the fact that the signal envelopes of deception jamming and target return are different. The algorithm can detect the existence or not of a deception jamming and find its initial time after calculating the difference between the signal's information dimension and the signal envelope's dimension. Simulation results show that this approach can detect deception jamming effectively when JNR is no less than 6 dB, and the detection probability is higher than 90%.

In order to meet the design requirements of high-precision and high-dynamics star simulators based on Liquid Crystal on Silicon (LCoS) splicing, the traditional splicing method and correction method were improved and optimized. A method of splitting a polarizing beam into three components was proposed to perform optical splicing. Based on the new splicing structure, a system with large papillary distance, low distortion, and an achromatic lens was designed, and the image quality was evaluated. The impact of aberration on the star position in dynamic star simulator was elaborated. A method to correct the position aberration based on the star simulator optical system was presented in details. Results show that, when an optical system is within a 12°×12° field of view, the relative distortion is less than 0.035%. In addition, when the Nyquist frequency is v= 60 lp/mm, the full field of MTF is no less than 0.8. The new splicing structure can improve the consistency of the contrast ratio between the two pieces of LCoS, and enhance the recognition rate of dynamic star simulator in a dynamic test of the star sensor. The new correction method can improve the accuracy of a star map display. The angular distance error is less than 7? after testing, which is in good agreement with the design requirements.

In order to adapt Karhunen-Loeve (KL) transform to multidimensional data, meanwhile to find out the optimal transformation of multidimensional vector matrix, a Multidimensional KL (MKL) transform based the theory of multidimensional vector matrix is proposed. First, the multidimensional covariance matrix is defined using the theory of multidimensional vector matrix, and the multidimensional characteristic matrix vector is solved. The, MKL transform is defined and the multidimensional data are mapped to the projective space. Simulation results show that the MKL transform in one-dimension has the same optimality as the KL transform. For 3D video data block, the average EPE can be as high as 99% by integral transformation, and MKL realizes decorrelation completely.

To track the Direction of Arrival (DOA) of the moving targets quickly and accurately, an adaptive subspace updating algorithm with a variable forgetting factor is proposed. First, this tracking algorithm adaptively adjusts the weights of current and historical data in a covariance matrix according to the DOA change speed. Then the maximum likelihood estimation algorithm is used and the Sequence Quadratic Program (SQP) is applied to optimize the likelihood function in order to reduce the computation cost of the algorithm. Experimental results show that the proposed DOA tracking algorithm has the ability to track coherent sources and obtain acceptable tracking results even under the condition of low SNR and small snapshot number in comparison with other methods.

The statistic features in Cntourlet domain are employed to build the natural statistic model and the tested image model first. Then a no-reference assessment algorithm in Contourlet domain (SCIQR) is proposed. Experiment results on subjective databases show that SCIQR outperforms the classical full-reference image quality assessment algorithm and the universal no-reference algorithm no-matter on single distortion type images and on the set of different types of distortion images. This demonstrates that SCIQR has good universality.

A time reversal focusing method based on transfer matrix estimation is proposed for far field targets. In this method, first, the least square estimation of the transfer matrix is obtained using detection and reflection signals of the targets, based on which the time reversal operator (TRO) is calculated. Second, the eigenvalues and corresponding eigenvectors are obtained by performing eigenvalue decomposition of TRO, where the number of targets can be confirmed by the number of the principal eigenvalues. Finally, the far target focusing can be realized through calculating the objective function of virtual observed points in the search area using noise vectors. Experimental simulations are performed and results show that the proposed method could realize time reversal focusing for far field targets effectively and accurately. The focusing performance of this method is better than that of the selective focusing method based on decomposition of TRO.

Because that the passive inductance is not easy to be integrated, a new ideal floating inductance based on Modify Differential Difference Current Conveyor (MDDCC) is proposed. The inductance values of the current conveyor in ideal and nonideal conditions are analyzed. This circuit consists of only two MDDCCs, two resistors and one grounded capacitor. So the structure of the floating inductance is simple, the inductance value can be adjusted independently by the two resistors and the capacitor. The circuit is used in continuous time voltage mode filter of two order low pass, and two order band pass and current mode high order filters. Computer simulation of PSPICE is carried out using parameters of the 0.18 μm CMOS process. Simulation results show that the proposed floating inductance has good circuit's work ability.

The composite orthogonal, composite coded, Discrete Cosine Transform (DCT), and Discrete Wavelet Transform (DWT) group fingerprinting schemes are constructed. Based on these schemes, the group traitor tracking algorithms (e.g. maximum correlation detector) are implemented. Experiments are carried out to compare the performance of these constructed group fingerprinting schemes. Results show that in all cases the group fingerprinting schemes can perform group detection effectively. The composite coded group fingerprinting scheme performs the best in noisy environment and after JPEG compression. The DWT scheme performs the best under group colluding attacks from different groups and the signal group.

Current Common Control Channel (CCC) allocation in Cognitive Radio (CR) Ad Hoc network possesses several shortcomings, such as poor allocation performance in high fluctuating topology structure and poor robustness to the primary activities. To overcome these shortcomings, a novel neighboring group based CCC allocation algorithm is proposed. First, the CR node and its 1-hop neighbors are grouped based the spectrum heterogeneity. Then, the mapping from the grouped CR nodes and the idle CCC set is formulated as a maximum edge biclique problem. Finally, a novel utility function for each available CCC selection is established according to the control massage delay and its throughput; and the CCC in the set with the largest utility function value is selected as the optimal CCC. The algorithm has the advantages of robustness to primary activities and facility in dedicating control channel allocation by the agile grouping of local nodes. Simulation results show that the proposed algorithm is effective in reducing the normalized end-to-end delay when CR node density and primary user number increase; and it adapts to spatial-temporal variation in spectrum availability.

An Enhanced Scaled Selection Combiner (ESSC) is proposed for Decode-and-Forward (DF) cooperative networks, where different modulation levels are adopted by the source and relay nodes to improve spectral efficiency. When the ESSC is used to combine signals from different diversity branches, the source-relay-destination branch is weighted by varying scale factors. As a result, the performance degradation caused by erroneous relaying in DF networks can be effectively mitigated by ESSC. Towards this end, the Bit-Error-Rate (BER) expression is derived for DF networks using ESSC under Rayleigh fading channels, and the scale factors of the ESSC are optimized to minimize the system BER. Theoretical analysis and simulation results demonstrate that a diversity order close to two can be achieved by ESSC. Furthermore, a power allocation scheme is proposed for DF networks based on ESSC, which can further improve the performance of cooperation.

As high dynamic ranges of background noise level in time, spatial and frequency domain seriously affect the performance of cooperative spectrum sensing algorithms, a novel cooperative spectrum sensing algorithm based on Normalized Spectrum (NS) is proposed. The local NS is uploaded to the fusion center, and the equal gain average of NS is treated as detection statistics to detect signals in the fusion center. It makes use of asymptotic normality and independence of Fourier transform to get the stochastic properties of Power Spectral Density (PSD). The mathematical expression for probabilities of false alarm is derived. In accordance with the Neyman-Pearson criteria, the closed-form expression of decision threshold is calculated. The simulation results show that the NS algorithm is robust to dynamic ranges of background noise level. With respect to Majority Voting Hard Decision (MVHD) cooperative spectrum sensing algorithm and Equal Gain Soft Decision (EGSD) cooperative spectrum sensing algorithm based on energy, the NS algorithm has better performance and broader applications.