For the coupling problems of d-axis and q-axis currents in the synchronous rotating frame, the diagonal matrix decoupling control method is proposed. The field current i_d and torque current i_q are decoupled into no-interfering variables by increasing a decoupling compensator in the motor current loop, so as to ensure the accuracy of PMSM's output torque. Furthermore, this method is compared with the feedforward decoupling control method and the feedback decoupling method. It is shown that the proposed method has better robustness and decoupling effect when the motor speed is high or the motor parameters change.

The vibration isolation performance of energy-regenerative semi-active suspension with variable stiffness and damping is investigated. A quarter car model with one degree of freedom with semi-active suspension model is used as the research subject. First, the magnitude frequency response characteristics of the model to road bumpiness excitation and external force excitation are analyzed. Then, a simple switch on-off damping control algorithm is designed based on the skyhook control method. Nine different on/off control modes are defined subsequently. Finally, with the input of sinusoidal function and random vibration, the control performances are simulated and compared using Matlab/Simulink. The results show that the control modes, which have on-off status for both variable dampers, lead the suspension to have excellent performance of vibration isolation.

In order to improve the performance of lane departure warning system, an unintentional lane departure analysis method is proposed. This method combines the driver's operation characteristics, vehicle's motion characteristics and the relationship between the vehicle and the lane. First, the unintentional lane departure is classified into two parts: lane departure by fatigue and lane departure by secondary task. Then, experiments of unintentional lane departure are carried out through the co-simulation platform based on CarSim and LabWIEW. Twelve drivers of different genders, proficiencies and driving behaviors are selected to participate the experiments. The unintentional lane departure parameters are collected and analyzed, including the driver's operation behavior, the motion characteristics of the vehicles and the relative motion position between the vehicle and the lane. Finally, an unintentional lane departure recognition model is constructed based on Gaussian Mixture-Hidden Markov Model (GM-HMM). The recognition results show good performance of the proposed model in online and offline tests.

IN order to control the aerodynamic force and noise of ground vehciels, unsteady flows in the wake of three-dimensional (3D) vehicles with different rear slant angles (0°, 25° and 35°) were experimentally studied with fluid dynamics, spectrum and correlation analysis for data processing, using hot wire anemometry. Two frequencies measured in the wake of 0° model are resulted from the unsteady flow emanating from the recirculation bubbles in the flow direction and the side direction respectively. Four frequencies detected for the 25° model are ascribed to C-pillar vortex, dynamics attachment flow on the rear slant, the structures emanating from recirculation bubbles in flow direction and resonance effect. Three frequencies captured for 35° model are connected to the structures emanating from the recirculation bubbles in flow direction and C-pillar vortex. The differences are significant among different models. Flow separation position, flow on the rear slant, C-pillar vortex are all different for different slant angle, which explains the physical meaning for each unsteady flow, the quantity and spatial distribution of dominant frequency.

“Pseudo subjective” method based on driving parameters of vehicles are commonly used to evaluate shift quality. This method indirectly reflects driver's subjective feelings by analyzing the vehicle running parameters, but ignores the human body's reaction in the shift process. In order to reflect the driver's subjective feelings, the driver model in the vehicle under shift condition is established and the human response spectrum is introduced to solve the problem for the first time. Gray relational theory is used to build the evaluation system. The system is verified by experiments to reflect the driver's feelings well.

The aerodynamic drag of five axes heavy duty vehicle with complex underbody structure was studied. The model with the real vehicle structure shape was established. Then, the macroscopic behavior of fluids in external flow field was simulated using the computational fluid dynamics software XFlow, which is based on Lattice Boltzmann Method. The air resistance coefficient of the whole vehicle model was obtained by lattice scale optimization test, and the influence of the underbody structure on the aerodynamic drag was investigated. The vehicle surface pressure distribution, the surface velocity field of the rotating wheels, the velocity distribution in the external flow field of the model, the turbulence intensity distribution, and the particle trajectories under the vehicle were analyzed. Results show that the drag coefficient error of the truck model is within 6%. Through the underbody structure optimization, the aerodynamic drag coefficient of the whole vehicle model can be reduced by 23%.

In order to understand the handling stability of tanker and explore effective stability control strategy, the liquid sloshing in partially-filled tanks with circular or elliptical cross-section is simulated using FLUENT, and the corresponding parameters are recorded. Then, the recorded data are processed in time and frequency domains. It is discovered that liquid sloshing presents periodic characteristics. The odd oscillation modes have asymmetrical waves, which contribute to the moving of the gravity center of liquid. Besides, the first oscillation mode contributes most to the moving of the gravity center. At the same time, the even oscillation modes have symmetrical waves, which do not contribute to the moving of the gravity center of liquid. It is also found that the filling percentage of liquid and external excitation are the main factors. The growth of the external excitation increases the nonlinear feature of liquid sloshing. Also, the growth of the liquid filling level also increases the nonlinear characteristics of liquid sloshing.

Video-processing technology was used to acquire detailed spatio-temporal data of pedestrians, and the underlying law of the conflict between violating pedestrians and turning vehicles was analyzed, then a model of pedestrians' crossing position distribution was proposed. First, on the basis of the survey data, the relationship between conflict and pedestrian crossing distribution was qualitatively analyzed. Then, based on the analysis of pedestrians' crossing positions at seven surveyed crosswalks, the impacts of crosswalk length and elapsed time on pedestrians' crossing position distribution were investigated. Further, a Gauss function was applied to model the pedestrians' crossing position distribution at different times. Finally, the formulas about elapsed time, crosswalk length and three model parameters were established. The adjusted R² of the fitting models were 0.9, 0.93 and 0.97 respectively, which indicates that the proposed models can well represent the law of pedestrians' crossing position distribution. The proposed model was verified and the role of the model on determining the conflict zone was illustrated. An optimization method to reduce the conflict zone was put forward.

In traditional hazardous material transportation model only the material transportation and only one time transportation were considered, but the special vehicle for hazardous material transportation was not considered. A hazardous material transportation scheduling model was proposed. In this model, it is assumed that the material needs to be transported for multiple times to meet the demand, the constraint condition is that special vehicles must be used to transport special hazardous materials, and the objective is to minimize the total transportation time. The multilayer coding genetic algorithm is applied to find the solution of this model. The correctness of the proposed model is verified by MATLAB simulation.

To reveal the competition and cooperation evolution mechanism between urban rail transit and traditional bus, the evolution model of the self-organization systems in urban passenger transport flow is constructed. On this basis, the competition model consisting of urban rail and traditional bus is established with the coexistence of both competitive effect and cooperative effect. Finally, as a case study, the evolution mechanism in different stages in Wuhan city is analyzed. The development feature of the competitive and cooperative effects of urban public transit at each stage is obtained and corresponding strategy is illustrated. This study provides a reference for the planning and development of urban rail transit and public bus networks.

An extended finite element model (XFEM) was constructed by ABAQUS to investigate the propagation of the traffic induced reflective cracking, and the anti-cracking characteristics and mechanism of geotextile. The influence factors of tensile modulus and placement location of the geotextile were analyzed. A repeated loading test on the material test system was conducted to validate the XFEM model. It is shown that the crack propagation induced by repeated loading can be classified into four stages. The larger the tensile modulus of the geotextile is, the slower the crack propagation will be, the stress intensity factor of the crack tip decreases and the internal force for crack propagation is insufficient as the crack propagates. The most cost-effective tensile modulus of geotextile is 100 MPa. Long-term force conditions of the geotextile will not change with the tensile modulus, and the geotextile plays a greater role in the early stage of crack propagation. Geotextile placed at a higher location, the crack length is bigger. Geotextile placed on the top of base course will absorb most concentrated stress to have the best cracking resistance.

The mechanical behavior of concrete specimens (cube, L=100 mm) subjected to a mixed path of stress and strain loading was investigated. The specimens were statically loaded into a true triaxial apparatus. First, the stresses in all three directions were simultaneously increased to a specified value p. Then, the strain in Y axis was increased monotonically while maintaining a constant Z-axial stress and a constant strain rate ratio between Y and X axes. The experimental results show that the compressive strengths of the concrete increase with the minimum principal stress, and decrease linearly with the increase in the strain rate ratio, which are higher than uniaxial compressive strength, and the maximum value is 3.4 times of the uniaxial compressive strength. The hydrostatic pressure does not affect the initial shear modulus when the concrete specimens experience only monotonic increase in hydrostatic pressure, and a coinciding part is found in the curves of shear stress versus shear strsin for different experimental groups. The peak value of volumetric strain increases with the minimum principal stress and the strain rate ratio. The volume of the concrete specimen decreases first and then increases, volumetric dilatancy phenomenon occurs. The peak strength and peak volumetric strain appear at the same time when the minimum principal stress is 10 MPa. Then as the minimum principal stress increases the time interval between the peak strength and peak volumetric strain increases, and the peak volumetric strain appears latter.

The bending performances of 21 bamboo plyboard enhanced glue-lumber beams were studied. Under the effects of bamboo plyboard layer number and thickness, the timber beam failure modes, ultimate load and mid-span deflection parameters were analyzed. Experimental results show that, compared to pure glue-lumber beams, the failure load of beams, with 1-3 layers of scuffing bamboo plyboard on the bottom, is increased by 16.8% ~ 45.9%. With the increase in the number of bamboo plyboard layers, the deformation capacity of the beams becomes stronger. However, the intercalated sliding becomes more obvious, which makes the bearing capacity of the beams decreases instead of increases. When the total thickness of the bamboo plyboard remains the same, the decrease in the thickness of each layer bamboo plyboard has no obvious influence on the failure modes of glue-lumber beam reinforced by bamboo plyboard. The total deformation is obvious and the ductility increases. The span-section strain of reinforced glue-lumber boards accords with the plane section assumption. The glue-lumber on the top of the beams reaches the compressive strength and has a certain plastic height when failure occurs. The scuffing-bamboo on the bottom of the beams reaches to the ultimate tensile strain. A formula to calculate the flexural capacity of the bamboo plyboard enhanced glue-lumber beams was put forward, which was verified. The optimum ratio of the thickness of the bamboo plyboard was given.

In order to accurately analyze the static and dynamic responses of steel-concrete composite beams, the reasonable element displacement function of the composite beam is derived based on the contact theory and finite element method. The stiffness equation of the composite beam element considering shear slip effect is deduced. Combined with classical dynamics theory, a method of calculating the natural frequencies and vibration mode of the composite beams without damping is established with consideration of the shear slip effect at the interface. As a case study, the deflection of a simply supported steel-composite beam under vertical load and distributed shear slip strain difference along the beam is analyzed. Results show that the finite element solution is in good agreement with the experiment results. Using stiffness matrix the natural frequencies and vibration modes of the beam considering the shear slip effect are calculated and compared with complete interaction results, which suggest that the effect of shear slip effect should be considered in dynamic response calculation. This study provides a reliable calculation method for practical engineering design.

In a heavy duty diesel engine equipped with regulated two-stage turbocharging system, the impacts of different turbocharge bypass valve opening on fuel consumption and emission are explored. Results show that, at low speed, opening the bypass valve helps to reduce NO_x emission, but in large load it causes the deterioration of the fuel economy. At high speed and low load, increasing the opening of the bypass valve helps to reduce pumping loss, the maximal drop in fuel consumption can reach 7.6%. While in middle and large load, due the effect of the interaction between pumping loss and indicated thermal efficiency, opening the bypass valve can reduce NO_x emission without worsening the fuel economy premise. At different speeds and low load condition increasing the opening of bypass valve, or at middle and large load condition closing or reducing the opening of the bypass valve can achieve the best fuel economy, without obvious deterioration of NO_x emission that the smoke emission is less than 4%. Under the optimal control strategy of turbocharged diesel engine, compared with the original single-stage turbocharged engine prototype, two-stage turbocharging in low speed has the advantage of high fuel economy.

The accuracy of wall functions needs improvement for near-wall flow in engine radiators. In order to make wall functions suitable for engine radiators, Improved Blended Wall Functions (IBWFs) were customized by blending the Standard Wall Functions (SWFs) and introducing influence coefficients of pressure gradients. Numerical simulations were conducted in Fluent. In the simulations, different near-wall treatments including IBWFs were applied on a tube-and-corrugated- fin type radiator. Experimental verifications were carried out in RWT800 radiator wind tunnel . The results indicate that the IBWFs save more than half of computational time of SWFs, meanwhile, the deviations of pressure drop and temperature differences between inlet and outlet decrease 9.1% and 14.5% respectively. The IBWFs provide a more efficient and reliable near-wall treatment for radiator numerical simulations.

To evaluate the combustion characteristics of a two-stroke glow-plug ignition miniature Internal Combustion Engine (ICE), a test platform was constructed and combustion diagnoses were implemented. Analogy analysis with conventional Gasoline Engines(GEs) shows that miniature ICE has lower Rate of Heat Release(ROHR), higher cyclic variations and longer combustion duration, about 20 °CA longer than those of conventional GEs. Besides, miniature ICE has lower indicated mean effective pressure (p_mi) and combustion pressure. Due to the unstable ignition of glowplug, the Start of Combustion (SOC) fluctuates considerably, leading to micro combustion instability. Transient analysis based on the captured first combustion cycle shows that high residual gas fraction of miniature ICE is the primary reason for the poor combustion characteristics of micro combustion. Without residual gas(the first combustion cycle), the ROHR curve of miniature ICE approaches those of conventional GEs and the combustion duration is approximately 22 °CA, nearly 50% shorter than those of stable combustion. Both engine block temperature and the residual gas affect the SOC and combustion duration, but the residual gas is the main factor. The more amount of residual gas and the higher the engine block temperature, the lower the p_mi because that both of them impair the increase of intake charge. At 150 ℃, the p_mi in steady combustion state is merely 0.17 MPa.

In order to solve the problems of low efficiency and high cost of the existing hydraulic transmission of wheel loader, a digital controlled hydraulic transmission (DHST) scheme of fixed step ratio automotive transmission was proposed based on digital hydraulic components and their control technology. To ensure the engine work steadily in the economic region, combined with the shovel operating condition of the wheel loader, fuzzy shift control strategies of the digital pumps and motors displacement were designed to realize the dynamic optimal matching between the combustion engine and the digital pumps, digital motors and the changing load. The feasibility of the scheme was verified by AMESim simulation. This research provides a new design concept for the wheel loader transmission system, and it has certain reference significance for the design of other mechanical transmission system.

To accurately solve the kinematic and mechanical parameters of the all terrain articulated tracked vehicle in the process of pitching movement, mathematical modeling method is adopted to establish the kinematic and mechanical models of the vehicle in pitching movement. The theoretical formulas of the maximum pitching angles, the maximum pitching heights, the motion trail radii of the centers of gravity of the front and rear vehicles, and the maximum pulling force of the articulated units are derived. Then, taking a specific articulated tracked vehicle as an example, the formulas are numerically solved. Finally, a virtual prototype of the articulated tracked vehicle is adopted to verify the theoretical models. It is shown that the theoretical and simulation results are in good agreement, the relative errors between the theoretical and simulation results do not exceed 15%. This error range meets the requirements of engineering practice. The results of this research work can provide a theoretical basis for designing and optimizing the structure of all terrain articulated tracked vehicles, and selecting the type of the pitching hydraulic cylinders.

Simulation analysis and experiments are performed to investigate the influence of the concentration of abrasive particles on the processing quality of abrasive flow in mesoscopic scale. Both analysis and experimental results show that with the increase in abrasive concentration, the dynamic pressure and turbulent kinetic energy increase, and the grinding effect is enhanced. The shapes of grinded workpieces are measured by zoom non-contact 3D measurement and scanned by SEM, and the results are similar to that of simulation. In other words, with the increase in abrasive concentration, the workpiece roughness decreases, the hole quality and workpiece performance are improved, the workpiece surface quality is enhanced.

According to the deformation of copper thin-wall part in micro milling, a mathematical model of the milling force was established. The method of quadrilateral thin plate element FEM was used to predict the deformation caused by the cutting force. Experiments with diamond coated cutters were carried out to the validate the feasibility of the model and the precision of the deformation prediction. Optical Microscope (OM) and Scanning Electron Microscope (SEM) were used to observe the milled surface. Results show that the proposed mathematical model and analysis method can effectively predict the milling deformation of copper thin-wall parts. There is a close relationship between the load position and the thin-wall deformation. The cutting deformation increases quickly with the decrease in the thickness of the thin wall.

An electrostatic torque feedback approach is used to solve the underdamping problem of a novel rotational gyro. The electrostatic torque feedback structure is designed, modeled and validated by finite element analysis. By comparing the dynamic responses of the gyro to different feedback conditions, the approach is demonstrated to be able to significantly improve the dynamic characteristics of the gyro. As feedback voltage increases, the system can infinitely approach the critically damped state from the underdamped state. Due to the restriction of dielectric breakdown and the side effect of the feedback strength on the magnitude-frequency response, the feedback voltage should be about 100 V, where, the overshooting of the unit step is reduced from 87% to about 5% and the settling time is reduced from 1.3 s to about 0.1 s compared with the open-loop system.

A turntable to be used in laser communication laboratory experiments needs positioning accuracy of ±1.5″. In order to check the technical requirements, a high accuracy angle measurement system, composed of regular polygon error, centering device, 0.2″ two-dimensional optoelectronic autocollimation and supporting, was set up. The collimation error of the system was theoretically analyzed, which shows that the off-center between the regular polygon mirror and the turntable contributes little impact on the positioning accuracy of the measurement results, which generally can be ignored. However, the adjustment error of the autocollimation has great influence on the measurement system. In order to eliminate the system error, on the basis of theoretical analysis, a new method based on data processing to compensate the adjustment error was proposed and verified by contrast experiments. Results show that a positioning accuracy σ=0.936″, was obtained by timing experiential fine adjustment, while the positioning accuracy σ″ = 0.922″ was obtained by the proposed new method. The relative error between σ″ and σ is -1.496%, which satisfies the technical requirement. Contrast experiments also demonstrate that the proposed method is convenient and reliable.

In view of new material 36MnVS4 with higher fatigue strength, the fracture splitting performance of the connecting rod was studied in comparison with C70S6, and reason of the fracture splitting defects was analyzed. In term of the brittle cleavage fracture feature of the new material mode I crack, the K criterion and the maximal principal stress failure criterion were determined. Meanwhile, the splitting Finite Element Model (FEM) of connecting rod was established with starting notch. By observing the stress concentration zone and the deformation distribution of the fracture splitting process, the crack initial position and time of the connecting rod and the plastic deformation of the notch and big-end in fracture surface were acquired. Considering the crack propagation velocity and the cracking load, the main defects were investigated and the fracture mass was evaluated. Results show that, under the same geometrical parameters and load, compared with the C70S6, the 36MnVS4 connecting rod production has more defects in the fracture splitting process because that the initial position of the crack is scattered, the crack propagation is instable, the notch sensitivity is high and there are more ferrites. However, the new material has a lower out-of-rounding as big-end hole and a smaller fracture surface plastic deformation after the fracture splitting.

The forming process of the front-end three-dimensional skin of China Electric Multiple Units (CEMU) was simulated and analyzed using the general finite element software Autoform. The influences of the process parameters such as blankholder force, drawbead, die radius, draw depth, friction coefficient and die gap on the forming of the V-section skin with large thickness, shallow drawing depth and twisted structure were investigated. The reasons for the forming defects such as insufficient stretching and poor contour accuracy were analyzed and the controlling methods were proposed. The results show that insufficient stretching is mainly originated from the lack of flow resistance, which can be controlled by setting proper drawbead and increasing the blankholder force; for the investigated long strip skin in this work, variable intensity drawbead is recommended. Poor contour accuracy is mainly caused by large springback. Process parameter optimization can improve the contour accuracy to some extent, and degrees of the influence parameters are in the order: die radius > friction coefficient > die gap > draw depth. Compensation to the surface based on springback can significantly control the contour accuracy. For the 4 mm 5083-H111 aluminum alloy, the optimal compensation is 1.2 times of the springback amount. Forming tests were conducted according to the simulation results, and high quality compensations were achieved, which prove the validity of the proposed defect controlling methods.

The forming principle of three-dimensional surface by flexible rolling and the control methods of two kind of typical surface forming are illustrated. The stress and strain states of the deformation zone are analyzed. The deformations and their relationship in the three principle directions are also analyzed. The forming process of two typical surfaces are simulated using finite element method. Simulation results show that the positive and negative spreads appear respectively when saddle and convex surfaces are formed. No matter of convex surface or saddle surface, the larger the relative elongation of longitudinal fibers is, and larger longitudinal deformation is. The flexible rolling forming experiments are conducted. The measurement and analysis results show that, compared with the target value, the maximum forming error is no more than 3.2 mm. Therefore, the quality of the forming surface is good and the springback is small.

A novel hot-stamping beam with ultra-high strength is introduced to improve both rollover safety and lightweight performance, which is based on the mechanical characteristics of the side beams and the excellent properties of hot-stamping material. Finite Element Method (FEM) analysis is used to calculate single cantilever beam with impact load on the middle position. Results show that the maximum deformation of the hot-stamping beam is about 74.9% lower the load bearing capacity is about 223.6% higher than that with equal-mass. This demonstrates that this novel beam has extremely crashworthy performance and potency of lightweight. The rollover crash simulation and test of the original school bus are conducted, which show that the simulation result is in good agreement with the test result, suggesting that the FEM analysis is accurate. The deformation, acceleration and structure weight surrogate are established based on Radial Basis Function (RBF) method. Then, multi-objective evolutionary algorithms are employed to obtain the optimal solution set. The results show that the deformation is decreased by 38.2%, the acceleration of survive room is decreased by 23.4% and the structure weight is reduced by 63.9%. It is proved that the hot-stamping steel beam performs far better in both rollover safety and lightweight performance compared to the original one.

Stellite-6, St6 in short, St6+5%WC, St6+5%WC+1%RE cobalt base self-fluxing powders were employed to enhance the H13 steel surface in CO₂ laser cladding process. The microstructure, elements distribution, phase composition etc. of the coatings were investigated using Optical Microscope (OM), Scanning Electron Microscope(SEM) and X-ray Diffraction (XRD).The micro-hardness and were-resisting property of the coatings were studied using micro-hardness tester and friction wear testing machine. The analysis shows that a metallurgical combination is formed between the laser cladding layer and the surface. All cladding coatings are composed of γ-Co and γ-Ni solid solutions as the matrix phase, Cr-Ni-Fe-C,(Mn, Cr)7C₃ and Cr₂₃C₆ as the wild phase. Furthermore, the phase compositions of St6+5%WC, St6+5%WC+1%RE also contain WC W₂C and SiC phases. The micro-hardness of the cladding is in the range 560~710 HV_0.2. Under the same testing conditions, the wear-resisting performance of the four materials is in the order: St6+5%WC+1%RE > St6+5%WC>St6>H13 steel.

In order to improve the transient and steady-state performance of brushed DC motor, a novel control system is developed on the basis of dual closed-loop control system. For the speed loop, considering the nonlinear characteristics of the friction torque and the un-measurable load torque, an Extended State Observer (ESO) based Triple-Step (TS) nonlinear control method is developed. First, the speed controller is designed using the TS method, which consists of the steady-state control, the feed-forward control related to the reference variation and the error feedback control. The structure of the designed nonlinear controller is concise. Then, an ESO based on model compensation is applied to estimate and compensate disturbances. The estimated errors are bounded and can converge to a prescribed range in a finite time. The estimated error is considered as disturbance input with the bounded amplitude. The method to analyze the robustness of this system is discussed in the framework of input state stability theory. For the current loop, current dynamics is neglected due to the fast electric response, the commonly used feedforward controller and PI feedback controller are designed to reduce torque ripple. The designed control system is evaluated by experiments. Results show that the proposed controller can improve the transient and steady-state performance of brushed DC motor significantly. The effectiveness and engineering implementation of this controller is also verified.

In view of the electromechanical-coupled vibration problem of electric wheel, the electromagnetic torque of permanent magnet synchronous motor is derived in taking account the effects of space harmonics of the permanent magnetic field. Then, based on the equations of coupled torsional and horizontal vibration, the inherent characteristics of the electric wheel system are analyzed. The electromechanical-coupled vibration model of the electric wheel is established by introducing closed-loop control of d/q axis current and combining the electromagnetic torque model and vibration equations. Based on this model the influence rules of electromechanical coupling effects on torque ripple and vibration performance are studied. Results show that the speed fluctuation of the rotor arising from torque ripple will in turn exert an effect of weakening on the torque ripple, especially at the rotating modal resonant operating condition of the electric wheel. The third modal damping ratio of the electric wheel coupling system is evidently influenced by the electromechanical coupling effects and varies with electromagnetic parameters, among which the magnet flux linkage has the largest impact.

In order to solve the problem of engine misfire fault, a diagnosis method based on cylinder pressure identified by Fourier Transform (FT) and L-M optimized BP neural network is proposed. The BP neural network is trained by the steady state simulation data from AMESim to obtain the relationship between open value of valve and frequency, and FT is used to identify the cylinder pressure. The misfire failure can be diagnosed by comparison of the FT identified pressure with the pressure mapped from crank speed. By offsetting the phase and frequency of the identified model, the accuracy and generalization of the identified model are improved. When misfire fault occurs, high tracking ability can be obtained by re-offsetting the phase and frequency of the identified model. Two random and independent open values of the valve are used to verify the proposed method, Results show that no matter engine at high speed with light load condition or at low speed with heavy load condition, this method can precisely diagnose the signal-cylinder continuous misfire fault and multi-cylinder random misfire fault.

In order to guarantee the security of event update messages in transmission, a novel protocol is proposed based on the group key agreement. With the utilization of hash-chain keys and discrete logarithm, the attackers and malicious sensor nodes could not forge or modify the messages. Simultaneously, as the wireless sensor network is composed of massive low-energy sensor nodes with limited computing capacity and short life period, it is vital to save the battery energy and reduce the consumption. Therefore, in the initialization phase, namely before message transmission, the key agreement is completed with help of the server. Theoretical analysis and experiment verify that the proposed protocol is better than other protocols both in security and efficiency.

This paper proposes an improved Greedy Randomized Adaptive Search Procedure (GPASP) algorithm, named GPHS (GRASP for Hitting Set). First, in the construction process, a score mechanism is proposed based on the dynamic degree of components namely d-covered to select one component which is the most likely component of min-length hitting set, avoiding to add the unlike component into the current solution. Then, in the local search process, a championship mechanism is designed based on the dynamic degree of components namely d-covered to remove the unlike redundant component from the current solution. In addition, a time complexity analysis of complete and incomplete algorithms is given. Experimental results show that, compared with a state-of-the-art complete algorithm, the proposed algorithm can find the optimal solution in a shorter time, and also find min-length hitting set with shorter length than the previous incomplete algorithm in the same test distances.

A Chaos Covariance Matrix Adaptation Evolution Strategy (Chaos-CMA-ES) optimization algorithm is proposed. The Chaos-CMA-ES algorithm uses the chaos operator to update the mean of the population on the basis of the original CMA-ES algorithm. Chaos-CMA-ES algorithm has good global search capability by using the improved operation. The comparative experimental results verify that the Chaos-CMA-ES algorithm has better optimization effect than other optimization algorithms for complex multimodal function. A case study of the network security situation prediction is examined to demonstrate the ability and applicability of the Chaos-CMA-ES algorithm. The prediction accuracy is higher than other methods.

On the basis of the pinhole model with radial and tangential distortions in computer vision, mathematical models of camera calibration based on computer vision and surveying adjustment fundamentals were established. Taking chess grid of Liquid Crystal Display (LCD) as reference object for camera calibration, Microsoft Visual Studio as the development platform, camera calibration software, Camera Calibration 1.0, was programmed based on MFC and OpenCV. Through calibration experiments for Canon 5D Marx II digital camera, and comparison with commercial software Lensphoto of close rang photogrammetry system, it was shown that the errors of calibration parameters were within ±1 pixel, and the correctness of the established camera calibration models was verified. The reliability of the Camera Calibration 1.0 software was validated. The precision evaluation of calibration parameters was achieved using the established calibration models, which overcomes the limit in conventional computer vision that only single precision for projecting errors of image point coordinates was evaluated. Furthermore, the method developed in this study was applied to the discontinuity investigation of the slope of Yangjiagou quarry in Changchun. The information of a large number of random discontinuities of rock mass slop was obtained to provide the full and accurate data for stability evaluation of the slop.

This paper presents an efficient parallel implementation of belief propagation and dynamic programming algorithms on an IBM CELL BE processor (Sony Playstation3) that speeds up stereo image matching. Overlapping windows and improved data cost with pyramid computation are adopted for implementation of belief propagation algorithms. Considering that real-world images are typically degraded by various types of noise, differing exposures and changes in light etc, Sobel edge and residual operator are used for image pre-processing to improve the stereo matching results. The proposed method has important application on hardware unification of automatic driving binocular computer vision together with multimedia visual technique.

To overcome the shortcomings of large amount of calculation and low efficiency of the traditional Otsu method in image segmentation, a GA-Otsu segmentation method for thermal waving images is proposed. The method gives full play to the genetic algorithm with global searching ability. The optimal threshold of the thermal waving image segmentation can be attained quickly and the segmentation time of the injury is shortened. Compared with the artificial threshold segmentation method, experiment results show that the proposed method not only keeps the basic shape of impact damage, but also the position of impact point segmentation is accurate. Finally, the feasibility of the proposed method is proved by the thermal wave image threshold segmentation.

In order to obtain high quality image information under the condition of compact optical lens structure, an approach for image restoration and reconstruction based on simple lenses computational imaging was proposed. First, by analyzing the characteristics of the Space Variant Point Spread Function (SVPSF) of the optical system, the method of SVPSF acquisition and measurement was presented. Then, the image sectioned restoration algorithm with SVPSF based adaptive sparse prior information was developed, which was used successfully in the restoration of blurry images through simple lenses. To validate the algorithm, the imaging results were obtained from through simple lenses imaging by the digital simulation and real images. Experimental results indicate that the algorithm can efficiently suppress the ringing effect and preserve the details of the images.

Aiming at Multiple Input Multiple Output (MIMO) radar system with large arrays, in which the number of arrays is comparable to the number of snapshots, a blind target detection method based on Linear Shrinkage-Standard Condition Number (LS-SCN) is proposed by combining the shrinkage algorithm of Covariance Matrix estimation and the large dimensional random matrix theory. By solving the optimization of the sample CM in the large dimensional regime and utilizing the M-P law, the relationship between the detection threshold and the shrinkage coefficient is derived. Single-target and multi-target detection algorithms based on LS-SCN are presented respectively. The method is not sensitive to noise changes and is suitable for large array system, which do not need to know the priori information of noise variance, target scattering matrix and target location. Simulation results show that, compared with SCN algorithm and Minimum Description Length (MDL) algorithm, the proposed methods significantly improve the performance of target detection under the circumstance that the numbers of arrays and snapshots grow at the same rate.

A novel precoding scheme is proposed with perfect Channel State Information at Transmitter (CSIT), which only takes one precoding and one Minimum Mean Square Error (MMSE) equalization at the transmitter and the receiver, respectively. Thus, the full diversity gain is obtained and the instant channel capacity approaches to the performance of Single-Carrier Frequency Domain Equalization (SC-FDE) in Gaussian channel. First, a theoretical mutual information bound between the transmitter and the receiver is presented based on the SC-FDE system with MMSE receiver. The, the precoding scheme is designed to achieve this bound. Through theoretical analysis, a closed-form expression on the outage probability of the system is derived. Simulation results demonstrate that the proposed scheme can reduce the outage probability of the traditional MMSE system by 2~3 dB and the instant channel capacity approaches to the performance of SC-FDE in Gaussian channel.

To achieve high-efficient data transmission, an efficient shoot-through framework architecture for network-on-chip is built, which makes trade-off between the centralized bus-based interconnection and the distributed NoC-based interconnection. Then, a torus-based hybrid network topology and shoot-through routing algorithm are proposed. The proposed framework breaks the inherent feature of conventional routing schemes that could remarkably reduce the length of routing path so as to improve the routing efficiency. Experimental results show that, compared with typical XY and congestion-aware adaptive routings, shoot-through framework with 10% more area overheads can improve the network saturation throughput by 84% and reduce the power consumption by 27.3% in average.

An algorithm for target threat assessment based on Back Propagation (BP) neural network optimized by Modified Particle Swarm Optimization (MPSO) is proposed to improve the prediction accuracy of target threat. In this MPSO algorithm, mutation operator and optimization for several parameters are introduced in PSO to avoid the particle plunging into the local optimization. The MPSO algorithm is employed to optimize the initial weights and thresholds of the BP neural network. Then the BP neural network optimized by MPSO is trained by training sets of different sample sizes. 60 sets of target threat data are adopted to test the performance of MPSO-BP in target threat prediction. Experimental results show that the prediction accuracy of target threat assessment algorithm based on MPSO-BP is higher than that based on some traditional algorithms, which proves the efficiency of the proposed algorithm in solving target threat assessment problem in spite of the small sample size of training set.

In this paper, application of signal interference technology to design a differential mode wideband bandpass filter and to suppress common mode noise is proposed. Due to symmetry and reciprocal property of this filter in common plane, symmetrical line is used as perfect electric wall for differential mode excitation and perfect magnetic wall for common mode excitation. To verify the presented concept, the prototype (mm) was designed and fabricated. The experimental results shows that the center frequency for differential mode bandpass filter is 7.85 GHz , the maximum return loss is less than -25dB ,3 dB fractional bandwidth is 61%, insertion loss is less than 0.2 dB and common mode rejection is lower than -20 dB for common mode (5.5~10.2 GHz). Good agreement can be observed between measured results and theoretical expectations.