In this paper, basalt fiber was used to improve the tensile properties to satisfy the requirement in automotive design. Meanwhile, Multi material design was applied in this case to achieve lightweight. The relationship between the fiber weight and the tensile properties was studied. Operation condition was obtained from reliability test in the proving ground using as constraints in Multi material design. The optimal process with combination of Multi-island Genetic Algorithm (MIGA) and Genetic Algoritm—Response Surface Method (GA-RSM) was used to search the optimal solution. The results show that basalt fiber has positive effect on improving the tensile properties. With the Multi material design and optimal process, the mass of the product is reduced by near 9%. It can be concluded that PLA composites is suitable for replacing the petroleum based materials in automotive and lightweight can be achieved with the combination of multi material design and MIGA.

In order to solve cause-and-effect problems during decision-making, in this paper, the Markov Decision Process (MDP) model is established by the analysis of car-following at first. Then, the state set and the action set are designed through the combination of driving simulator experimental data and driving risk principle. Third, the reward functions are designed according to different driving states. Finally, a causal reasoning mechanism during the process of decision-making is proposed and reinforcement Q-learning algorithm is applied to solve the MDP model. The feasibility and effectiveness of the proposed method are verified through the simulation tests with random driving conditions.

To get better response performance of a road feeling simulation actuator, a current control (Triple-step saturation prediction current control, TSPCC) algorithm is designed using the triple-step nonlinear method to solve the problems of the inherent delay of digital motor control and the output saturation of the inverter. The objectives are to improve the current response speed and reduce the current oscillation of permanent magnet synchronous motor of the road feeling simulation system. The triple-step nonlinear method has the advantages of being less sensitive to parameters and easy to implement in engineering. By transforming the inverter output saturation into model-based dynamic saturation limit, the dynamic feedforward control law and error feedback control law can obtain the optimal current response speed. The next step predicted current value is used to reduce overshoot and oscillation of the current due to delay. Simulation and experimental results demonstrate that the proposed algorithm can significantly increase the response speed of the current and reduce the current fluctuations in the steady state, and is suitable for smooth touch feeling and real-time road feeling.

The traditional CFD simulation method generally only considers the effect of wind load on the aerodynamic performance of the automobile and neglects the influence of the coupling between the vibration of the body structure and the airflow, resulting in some discrepancies between the calculated results and the real car driving conditions. Taking the 1/4 standard MIRA model as the research object, the fluid-structure interaction effect is introduced into the numerical simulation through the bidirectional explicit fluid-solid coupling simulation method, and the aerodynamic forces, surface pressures, vibration frequencies and body attitude angles under different conditions are obtained. The differences between the current simulation results and the traditional simulation methods are analyzed. The accuracy of the current simulation results is verified by the wind tunnel test. Comparing with and without coupling simulation, the experimental results show that the coupling simulation is more consistent with the experimental results and the deviation of the data is within 5%, which verifies the accuracy of the coupled simulation method. The fluid-solid coupling effect is more affected with the increase in vehicle speed, especially the impact of aerodynamic lift directly affects the vehicle handling stability. As a result, the effect of fluid-solid interaction cannot be ignored at high speeds.

To solve the problem of intelligent vehicle forward multi-sensor multi-target tracking and fusion, an algorithm of vehicle multi-sensor target tracking and fusion based on modified joint probabilistic data association is proposed. First, according to the relative motion of the vehicle coordinate system and sensor coordinate system, the multi-sensor data is transformed. Then, the single-sensor multi-target tracking based on modified joint probabilistic data association, the multi-sensor track correlation based on correlation sequential track association and convex combination fusion are adopted to achieve stable tracking and accurate fusion of the target. Finally, the experimental vehicle equipped with millimeter-wave radar and camera is tested in actual traffic environment. The results show that the target is tracked steadily and the fusion results have good accuracy which verify the feasibility and effectiveness of the proposed algorithm.

In order to solve the path planning problem of self-driving bus in the structured road environment, an improved path planning algorithm, named Regional-Sampling Rapidly-exploring Random Tree (RS-RRT) algorithm, was proposed for obstacle avoidance conditions. In the sampling phase, Gaussian distribution sampling and local biasing sampling were integrated to improve the search efficiency of the path planning algorithm. In the expansion phase of the random tree, considering the actual size of bus and obstacles, the Separating Axis Theorem (SAT) was used to detect the collision of bus and surrounding obstacles in real time. In the post-processing stage, considering the goal of safety and comfort，the driver's driving consensus, the safety distance model and path smoothing algorithm were combined to correct the planning path. In order to verify the effectiveness of the RS-RRT algorithm, the hardware-in-the-loop test bench of electro-hydraulic steering system for commercial vehicle was built. The simulation scenario was built by TruckSim, and the proposed algorithm was verified by the co-simulation software of MATLAB and TruckSim. The results show that compared with basic RRT and Goal-biasing RRT, the proposed RS-RRT algorithm has advantages in terms of number of nodes, path length and running time. The generated path can meet the dynamics and path tracking requirements of the bus.

For the shift actuator control during AMT engagement, an adaptive intelligent control strategy was proposed. The goal is to make the synchronizer displacement follow the target displacement trajectory closely by dual closed-loop control (shift motor current and synchronizer displacement). Considering the shift actuator parameter uncertainty and dynamic disturbance, a compensator control was added to get access to the actual shift motor feature. The compensator gains were obtained by the self-learning algorithm in neuron network, and the input of the self-learning algorithm is the error between shift motor actual current and predicted current by neuron network. Meanwhile, fuzzy adaptive control was utilized to regulate the parameters of PI control in the closed-loop of synchronizer displacement. The target synchronizer displacement trajectory was updated in time by offline and online self-learning strategy. The simulation results verify that compared to normal PID control, the proposed strategy has higher accuracy, better stability and faster response in following target trajectory.

Taking a lithium pouch battery of a hybrid electric vehicle as the research object, the influences of temperature and State of Charge (SOC) on battery capacity, DC internal resistance and open-circuit voltage were experimentally studied. Also the factors influencing the battery life were analyzed. The performance parameters of a large number of lithium batteries were statistically tested, the consistency and correlation of the battery performance parameters were investigated. Experimental results show that over high or over low temperature could lead to an increase in battery resistance and a decrease in battery capacity, which reduces battery life. Excessively low SOC can also cause the increase of internal resistance of the battery and the drastic change of open-circuit voltage, which not only reduces the battery life, but also weakens the efficiency of the battery. There is a certain correlation between the performance parameters of the batch battery, and its consistency is greatly influenced by the number of batteries. The optimization of the number of batteries, the use of temperature and the SOC not only improves the dynamic performance, but also significantly improves the service life of the hybrid electric vehicle.

To accurately estimate the saturation flow rates of approaching lanes at signalized intersections, the headways obtained in peak hours are expressed by time series. The steady state test method is proposed based on PP test. The abnormal headways of the approaching lane during green time are deleted to obtain the saturation headways. The saturation flow rate is estimated using the average value of saturation headways. The three through lanes of a typical signalized intersection in Changchun city are taken as examples to validate the proposed method, and the method is compared with classical Highway Capacity Manual (HCM) method. The results show that the proposed method in this paper can estimate the saturation headways accurately; the estimated saturation flow rate ranges from 1700 pcu/h to 1900 pcu/h, which are higher than that of HCM method.

Because the characteristics of many monitoring points, large data and low degree of data fusion of the gearbox in a rail vehicle, an anomaly detection method of the gearbox is proposed based on the correlation function fusion algorithm and fuzzy C means clustering. The correlation function fusion algorithm is employed to fuse the vibration signals collected from many monitoring points into a signal which can fully reflect the running state of the gearbox. The ensemble empirical mode decomposition (EEMD) method is applied to decompose the fused signal, and the energy entropy and singular entropy of the first n intrinsic mode functions (IMFs) are calculated. The fuzzy C means (FCM) clustering algorithm is employed to cluster the feature set to determine several abnormal situations in the gearbox. Through collection and analysis of the data during actual operation, the effectiveness of the proposed method is verified.

In order to meet the charging demand of electric vehicle for commute users near their workplace, a charging station location and capacity model is constructed. The objectives are to minimize the cost of building charging station and minimize the generalized charging cost of users. The convenience factor is introduced to quantify the charging expectation of the user to obtain the number of electric vehicles and the power demand of the charging station. The calculation example shows that the location of charging station is not sensitive to the threshold of convenience coefficient, but the economy of construction becomes worse with the decrease in the threshold of convenience coefficient. Therefore, the range of the threshold value of convenience coefficient under different programming objectives is suggested.

Based on the original path correlation model, considering the discrete vehicle, block and spatial characteristics of local traffic flow, path correlation model between adjacent crosses is established. This model compensates the shortcoming of the original model, which only considers a single index. Based on the calculation results of the path correlation model, signal control intersection groups are dynamically divided using the hierarchical clustering method. Case analysis and simulation results show that the new model of intersection group dynamic division based on signal control protocol can improve the regional Regulation efficiency.

To better understand the characteristics of public transport commute passengers with different categories and further to better meet their customized travel demands, it is necessary to find ways to identify public transport commuters’ stability accurately. Based on revealed preference survey data, public transport smart card transaction and network data, the travel chain of public transport commuter was obtained by data processing and matching. According to public transport travel behavior data of one month, five characteristic indexes were extracted to represent travel stability from the aspects of activity point, travel space and travel time. These feature indexes include the number of non-home activity categories, the proportion of typical travel chain, space balance degree, time stability and time concentration degree. Through applying FP-growth algorithm, the association rules among each characteristics attributes for different item-sets were mined depending on parameters of support, confidence and lift degree. Then, three categories of public transport commute passengers with apparently different stabilities were identified. Finally, the rationality of the proposed identification method was verified. The study results contribute to developing customized and targeted supply and demand management strategy for public transport travel, which would further help to improve the efficient and delicacy service level of public transport.

From the perspective of the transshipment delay risks of multimodal coordinated freight transport in nodes, this paper analyzes the coordination and optimization of multimodal coordinated freight vehicle allocation from the perspective of transshipment delay risks in transport node. With the comprehensive analysis of the risk factors of transshipment in multimodal coordinated freight transport node, the entropy weight method is used to evaluate the transshipment delay risks. The optimization model of multimodal coordinated freight transport vehicle allocation is built on the basis of considering the transshipment delay risks, and the NSGA-II algorithm is proposed to solve the optimization model. Finally, a computation example is given to analyze the effect of the allocation optimization of multimodal coordinated freight transport based on the transshipment delay risks.

In order to consummate the in-plane stability theory of self-anchored suspension bridge, firstly it is preliminarily determined that the bifurcation failure mode of this type bridge can not occur through the concept of structural stability. Furthermore, by introducing the displacement interference which is assumed as power series form in the deflection equation of the self-anchored suspension bridge, it is proved that the elastic bifurcation instability will not occur. Finally, a practical bridge is taken as an example, the numerical models according to the elastic stability and elastic-plastic stability theory are calculated and analyzed to obtain the load coefficient, bridge failure modes etc.. The results show that there is no in-plane bifurcation failure mode for self-anchored suspension bridge in the scope of application of deflection theory. Numerical analysis shows that the bifurcation instability in the actual bridge deck is caused by the breakage of the sling, which exceeds the application range of deflection theory, and the ultimate bearing capacity satisfies the safety requirements.

In order to study the mechanical properties of laminated slab with shear keys, the necessity of setting shear keys was analyzed based on the experiment and finite element simulation method by comparing the load-deflection curves of the laminated slabs with different number of shear keys. The stress characteristics of the laminated slabs were analyzed by comparing the loading process of the laminated slab with shear keys and the cast-in-place slab. The damage process and the shear force distribution law of the shear keys were analyzed respectively by extracting the shear force of the shear keys, and the simplified calculation equation for the shear force was established. The results show that the shear keys can significantly improve the stiffness and the bearing capacity of the laminated slab. Compared with the cast-in-place slab, when the concrete at the bottom of the laminated slab is cracked and the steel bar is yielded, a significant difference exists in the distribution law of the first principal stress of the concrete at the bottom of the slab, while a similarity occurs in the stress distribution law of the steel bars and the displacement distribution law at the bottom of the slab. With the increase in the load, the shear force of each shear key in the laminated slab increases linearly and then drops suddenly. The shear force of the shear key is bigger as the shear key gets closer to the bearing. At last, the layout method of shear keys is put forward.

This paper investigates the influence factors of the splitting strength of asphalt mixtures by means of numerical simulation. A micromechanical model for the splitting test of asphalt mixtures was established based on the three-dimensional (3D) Discrete Element Method (DEM) and the Cohesive Zone Model (CZM), and the splitting strength of asphalt mixtures with different aggregate gradations at different temperatures was simulated. Simulation results were verified by laboratory splitting tests. Results show that the splitting strength of asphalt mixtures can be simulated well using a micromechanical model based on CZM and 3D DEM. The aggregate gradation and temperature significantly affect the splitting strength of asphalt mixtures. At the same temperature, the splitting strength of asphalt mixtures increases with the nominal maximum aggregate size or when aggregate gradation becomes coarse. With the same nominal maximum aggregate size, the splitting strength of asphalt mixtures decreases with the increase in temperature.

In order to study the mechanical properties of silty subgrade and the silt subgrade slope stability under the freeze-thaw cycles, the static triaxial test was conducted with different water contents to study the variation law of the cohesion and the internal friction angle of silty sand after different numbers of freeze-thaw cycles in seasonally frozen area. The influences of freeze-thaw cycle on silt subgrade slope stability coefficient of the different slopes and different fill heights were analyzed by the finite element numerical model based on the test results. The results show that the influence of freeze-thaw cycle on cohesion of silty sand is gradually enhanced with the increase in water content, the cohesion can be reduced 70%-80% when the water content is large. The internal friction angle decreases firstly and then increases slightly with the increase in the number of freeze-thaw cycles, the influence of freeze-thaw cycles on the internal friction angle of silty sand is weakened with the increase in water content. The silt subgrade slope stability coefficient is decreased by the freeze-thaw cycles and the largest decrease after freeze-thaw cycles can reach 49% of stability coefficient without freeze-thaw cycles. The maximum fill height of silt subgrade should not exceed 4 meters in seasonally frozen area. If the subgrade fill height is less than 3 meters, the limit slope should not be steeper than 1∶1.5. If the fill height is more than 3 meters, the limit slope should not exceed 1∶2.5.

In order to study the mechanism of internal flow of large-scale hydrodynamic coupling under dynamic speed regulation, the dynamic speed regulation process is defined by UDF program in this paper, and a numerical method for calculating the transient gas-liquid two phase flow with a large-scale hydrodynamic coupling is established. The dynamic circulation flow of gas-liquid two phase fluid between the pump and the turbine wheel is produced, and the gas-liquid two phase distribution state is obtained. The results show that with a smaller slip ratio at the initial stage of speed regulation with the shielding effect of high energy fluid in the vane free region, the torque coefficient is significantly reduced. With the decrease in speed ratio, the turbulent kinetic energy increases gradually, and the turbulent intensity of the pump impeller is much larger than that of the turbine. In the inlet of the pump wheel, the turbulent kinetic energy has an extreme value, and the phenomenon is especially evident in the braking condition. The PIV test shows that the numerical simulation results are in good agreement with the test. The work of this paper provides the necessary theoretical basis for optimizing the transient performance under the dynamic speed regulation mode of the large-scale hydrodynamic coupling.

Due to the existing uncertainties caused by high non-linearity, friction, leakage and other external factors in the flight simulator with continuous rotary electro-hydraulic servo motor, it is difficult to accurately establish the mathematical model and also the traditional control strategies can not satisfy the high performance requirements. To overcome these problems, a compound control strategy composed of Predictive Function Control and Sliding Model Control (PFC-SMC) was proposed. In the electro-hydraulic servo system, the control principle of internal model control with state feedback is adopted, the internal model control principle and sliding model variable structure control is combined. The sliding model variable structure control can overcome external disturbing and parameters’ time-varying property. The internal model controller is designed using the predictive function theory, which can greatly suppress chattering of sliding mode structure, and achieve high precision and high frequency responding on tracking performance. The simulation results show that PFC-SMC algorithm improves the tracking performance and anti-interference ability of the electro-hydraulic position servo system greatly comparing with the traditional PID control, and effectively expands the bandwidth. The accurate control of electro-hydraulic position servo system of hydraulic simulation turntable can be realized.

Based on ergonomics, this paper designs and optimizes a multi-functional training machine structure for sitting-to-standing (STS) exercise and rehabilitation, which enables patients to carry out weight-loss rehabilitation training according to the track of the healthy body. The structure of lower extremity orthodontic exoskeleton and the top dumbbell spiral winding wheel were designed. The mechanical analysis and dimension optimization of the whole mechanical structure and main components were carried out. Users with lower extremity dyskinesia can get STS exercise assistance, and perform lower extremity muscle rehabilitation training according to the correct exercise plan of healthy people when they stand up, that avoids muscular abnormal exertion caused by the distortion of standing posture, thus obtaining safe and healthy rehabilitation.

In order to explore the mechanism of workpiece material removal during abrasive flow machining, the molecular dynamics method was used to simulate the machining process that SiC particles micro-cut single crystal Fe. The microscopic mechanism of action that affects the cutting force variation and the reasons for the continuous fluctuation of cutting force in the process are studied. The internal reasons of physical change of workpiece material and the working law of abrasive grains on workpiece are revealed under the micro-cutting process. Through the analysis of atomic displacement, it can be seen that the plastic deformation of the material in the micro-cutting process is mainly caused by the slip region generated in the sliding surface of the crystal lattice to form dislocations, and the atoms move along dislocation lines. At the same time, the accumulation of atoms and the mechanism of chip formation are discussed. The polyhedron template matching method was used to analyze the arrangement of workpiece atoms in the micro-cutting process, and the variation of the lattice structure in the workpiece material was discussed. The research can provide theoretical basis and technical support for processing single crystal Fe with abrasive flow machining.

The martensitic transformation of surface white layer in high speed hard cutting is studied in this paper. First, a martensite critical phase transformation temperature model was developed based on the phase transformation mechanism of white layer. Then the theoretical martensite fraction of the white layer was calculated combined with the M _S model. The calculated martensite fraction was compared with the martensite fraction measured by experiments. The results show that the value of the calculated martensite fraction is higher than that of the experimental value. The formation of martensite of the white layer is influenced by cutting heat and plastic deformation. The influences of cutting speed and flank wear on the martensite fraction are analyzed combined with the transmission electron microscope experiments.

Graded thickness and multi-corner designs are two efficient strategies to improve the energy dissipation of thin-walled tubes. In this study, the mechanical properties of multi-corner structures (square and nonconvex multi-corner tubes) with double surface gradients subjected to axial load were theoretically and numerically studied. An analytical formulation for the mean load of these types of novel structures was derived using the super folding element theory. The theoretical predictions agreed well with the simulation results obtained using the ANSYS/LS-DYNA software. The results show that increasing the corner number and thickness gradient can improve the energy-absorption efficiency. The combination of nonconvex multi-corner structures and double-surface gradients can lead to the increase in the specific energy absorption as high as 148% to 205% when compared to a square tube with a uniform thickness. This fully demonstrates the high efficiency of this type of combination in improving the crashworthiness of thin-walled structures.

In order to obtain a higher pressure capacity, the camber-type and plane-type double-layered split ultrahigh pressure dies were investigated. The stress distributions and pressure capacities of the two types of ultrahigh pressure dies were studied by finite element method. The results show that the novel plane-type double-layered split ultrahigh pressure die can effectively decrease the stress levels of the cylinder and first layer supporting ring. The pressure capacity of ultrahigh pressure die is significantly improved accordingly. Compared with the camber-type double-layered split ultrahigh pressure die, the Mises stress and maximum shear stress on the cylinder of plane-type double-layered split ultrahigh pressure die are decreased by 27.2% and 33.2%, respectively. Meanwhile, the Mises stress and maximum shear stress of the first layer supporting ring are decreased by 14.8% and 4.7%, respectively. The pressure capacities of the plane-type double-layered split ultrahigh pressure die are increase by 35.6% and 47.1% according to the maximum distortion energy criterion and maximum shear stress criterion. The experimental results show that novel plane-type double-layered split ultrahigh pressure die has a remarkably higher pressure capacity.

In this paper, ultrasonic welding technology was applied for welding the Cu/Al dissimilar metals lap joints. The peak temperature, microstructure and mechanical properties of the joints were analyzed by UT-325 thermocouple thermometer, Optical Microscope (OM), Scanning Electron Microscopy (SEM) and universal testing machine. Results show that the stress at the Cu/Al interface increases constantly with the growing of welding energy, which causes the increase in interfacial friction and heat generation. Furthermore, the interface is improved gradually thanks to the increasing atomic activity and the plastic flow capacity. In addition, swirling deformation is formed at the interface due to the uneven distribution of stress. The driving force at the interface increases with the growing of welding energy that can cause a more obvious uneven distribution of the stress and a more dramatic swirling deformation. This kind of deformation will cause the mechanical interlock between Cu/Al dissimilar metals which can increase the shear force of the joint. The maximum shear force of the joint reaches 3.72 KN. Moreover, intermetallic layer with the dominant phase of Al2Cu is found in joints with the welding energy of 1700 J or even higher. Fracture mode of joints is the mixed-mode of ductile and brittle fractures.

Taking the front floor of a car as an example, the finite element software Autoform was used for numerical simulation analysis of forming process. The defects such as splitting, wrinkling, and failure during the forming process of the sheet metal were predicted through simulation analysis. By adjusting the size of the blank holder force, the drawbead parameters, the offset between the punch and the die, and the friction coefficient, the simulation results were optimized to obtain a set of the most suitable parameters for actual production process. After setting the variable strength double drawbeads, the optimum conditions for the process parameters are: blank holder force is 650 kN, forming force is 5200 kN, the offset between punch and die is 0.8 mm and the friction coefficient is 0.15. Based on the simulation results, a forming experiment was carried out and a qualified forming part was obtained, which effectively shortened the development cycle of the die and reduced the design cost.

304L stainless steel was prepared by selective laser melting (SLM) technology. The microstructure, morphology and mechanical properties of as-fabricated parts were characterized. The results show that no obvious pores and inclusion defects are found in the metallographic structure. The surface roughness of the sample decreases with the increase in the inclination angle during the SLM process. To keep the geometrical characteristics of as-fabricated SLM parts, the thickness of thin-walled specimen should be more than 0.3 mm; the diameter of circular hole without support should be controlled in the range of 1 mm to 12.5 mm; the designed outer sharp angles of the specimen should be more than 5°; the inner sharp angles should be more than 15° when placed vertically to the base plate, and more than 8° when placed horizontally. It is found that the orientation of the specimen axis with respect to the working platform does not influence the tensile strength and yield strength. When the build direction of the specimens is 0° with the base plate, the maximum tensile strength of the specimens reaches 684.7 MPa. When the build direction was 30° with the base plate, the maximum post-fracture elongation of the specimens reaches 47.7%. Meanwhile, the impact energy has the maximum value of 196.3J, and the fracture mechanism is ductile fracture.

In order to study the influence of strain rate effect on the simulation precision of fiber reinforced composites, the Dynamic Enhancement Factor (DIF) is introduced to modify the modulus and strength of the composite. The modified Hashin failure criterion is used to establish a three-dimensional progressive damage constitutive model of composite materials, which considers the strain rate effect. The model is embedded into the Abaqus software by VUMAT subroutine, and the simulation of the projectile penetrating the fiberglass reinforced composite laminates is carried out. The simulation results are compared with the experimental results and the simulation results of the constitutive model without considering the strain rate effect. The results show that the constitutive model considering the strain rate effect can accurately simulate the process of projectile penetrating the laminate. Compared with the simulation results without strain rate effect, the accuracy is increased by 14.1% and 22.7% respectively by the proposed model, at the penetration speed of projectile at 1.8 m/s and 3 m/s. In the simulation of the constitutive model without considering the strain rate effect, the error increases from 28.7% to 36.9% as the penetration speed increases from 1.8 m/s to 3 m/s, however, the simulation error of constitutive model considering strain rate effect is relatively stable.

In order to investigate the friction mechanism of the bioinspired materials with integrated soft and hard layers, the mathematical model of the friction process of bioinspired soft-hard integrated materials is established based on the theory of adhesive friction, taking into account the structure of soft and hard layers. The friction process of bioinspired soft-hard integrated materials was analyzed by numerical analysis software “Abaqus”. The effect of material properties, such as the volume ratio and friction coefficient of soft and hard materials, on the friction coefficient of the bioinspired materials is discussed. Bioinspired materials with integrated soft and hard layers were prepared by 3D printing and friction experiments were carried out. Experimental results verify the theoretical analysis and simulation results.

The effects of creep aging and artificial aging on the fatigue crack propagation properties of 2524 aluminum alloy were investigated. The influence of aging conditions on the fatigue properties of the alloy was analyzed based on the microstructure evolution of the alloy under TEM. The results show that the microstructures of artificial aging and creep aging alloys are mainly Cu-Mg clusters and GPB zones after aging for 4 h, showing good fatigue resistance. Compared with 4 h, after aging for 9 h the microstructures of the alloy are mainly acicular S′ phase. The appearance of these coarse S′ phases changes the slip mode of dislocation, which reduces the reversibility of slip and accelerates the accumulation of fatigue damage. Near the threshold section of \mathrm{\Delta } K \le \mathrm{7} \text{?} M P a · m^{\frac{\mathrm{1}}{\mathrm{2}}} , the alloy fatigue resistance decreases significantly. The precipitate free zone appears at the grain boundary of the alloy after aging for 24 h, and the precipitates in the grain are also larger in size, further accelerating the fatigue crack growth rate of the alloy. Creep stress accelerates the precipitation process of alloy and compared with artificial aging, the creep aging alloy has larger precipitate size, higher yield strength and hardness, but lower fatigue resistance at the same aging time.

With the development of the Vehicle-to-Microgrid (V2M) technique, Electrical Vehicles (EVs), which are considered as the controllable loads and also as the distributed energy storage units, are used to alleviate the frequency fluctuation of microgrid caused by the intermittent new energy. This paper studies the load frequency control problem of microgrid with EVs. Frequency regulation characteristics of the diesel generator and EVs are analyzed, respectively. Then, the model predictive control strategy is proposed for the load frequency control of microgrid with EVs. The simulation results show that when the system is subjected to a step load fluctuation, the proposed model predictive control strategy can reasonably allocate the output of each system, and quickly eliminate the frequency fluctuation. Moreover, it also extends the life of the generators while providing high-quality power to customers.

In this paper, a novel POI recommendation algorithm based on clustering and matric factorization is proposed. The matrix factorization method is used to quantify the user's number of check-in for unknown location. The clustering method is used to divide the users into groups. In addition, the improved similarity calculation method is used to calculate the similarity among users. The experiment results on Yelp dataset show that the proposed method can improve the location recommendation recall rate and accuracy rate, demonstrating the better performance of our method compared to other methods.

A pedestrian-vehicle detection network (PVDNet) is presented for pedestrian and vehicle detection in driving environment based on deep learning. First, on the low layers, an improved skip connection called Multi-Level Skip Connection (MLSC) is proposed to accelerate the convergence speed and the accuracy of the model. Second, on the top layers, a Multi-Layer Features Fusion (MLFF) method is designed to improve the detection accuracy by combining the low-level features with the high-level features. Finally, on the output layer, an One-Dimensional Convolution (ODC) Method is proposed to reduce the model parameters and improve the detection speed by replacing the fully connection layer. Experiments of the proposed PVDNet were carried out on the PascalVOC2007, PascalVOC2012, MS COCO, KITTI datasets. results show that, compared with the original Faster R-CNN, the mean average detection accuracies on the PascalVOC2007, PascalVOC2012, MS COCO, KITTI datasets are promoted 3.7%, 6.1%, 5.6%, 9.62% respectively by using PVDNet.

In order to improve the recognition rate of scene classification and make it better used in target detection and behavior detection, an improved scene classification method is proposed. Firstly, the method uses naturalness in the Spatial Envelope to make the basic division of the scenes. Then, the visual feature channel model is applied to further classify the scenes in more details, resulting in the final semantic categories. Simulation results show that the improved scene classification method for semantic classification of scenes is superior to traditional scene classification method and has a higher recognition rate compared with other scene classification method.

This paper puts forward a digital signature based on discrete logarithm and hash function. After security analyzing, the proposal has the resistance to IEAs based on discrete logarithm and hash function. After performance analysis, the memory occupancy of wireless sensor nodes was effectively reduced by the signature scheme presented in this paper.

In the test of aircraft load parameters, some structures will produce large strain concentration points. In order to solve the problem of large-scale strain monitoring of aircraft structure，a temperature-sensitized, strain-reducing fiber grating large-range strain sensor is proposed. The fiber Bragg grating sensing theory is analyzed, and a new sensor base structure is designed and analyzed by finite element method. The temperature-calibration, temperature repeatability, and strain calibration tests were performed on the packaged sensor. The results show that in the environment of 10~60 °C, the packaged fiber grating strain sensor has a temperature sensitivity of 44.959 pm/°C, which is 4.5 times stronger than bare fiber and has a linearity of 0.999. The central wavelength change at the same temperature is ±4 pm. Under the condition of 0~2000 με, the strain sensitivity is 0.79 pm/με, which is 1.52 times lower than that of bare fiber, the linearity is 0.999, and the repeatability of multiple strain cycles is ±5 pm. The packaged sensor has good sensitivity and consistency, meets the accurate measurement of the aircraft's large-scale strain, and has a promising future in the monitoring of aircraft strain concentration structure.

The DV-Hop algorithm has low positioning accuracy and the Received Signal Strength Indicator (RSSI) algorithm is greatly affected by environmental factors, which lead to the generation of estimation error. To overcome these weaknesses and improve positioning accuracy and robustness, an improved algorithm combining DV-HOP and RSSI (DV-HOP-RSSI) is presented in this paper. The core ideas of the DV-HOP-RSSI algorithm are as follows. First, the single hidden layer feedforward neural network is used to obtain the trained network model instead of network topology, to achieve complete topology of the training set. Then the distance between unknown nodes and anchor nodes is calculated using the DV-HOP-RSSI algorithm. Furthermore, the location information of unknown nodes can be obtained by entering this distance into a trained network model. The simulation experiments were carried out in MATLAB, and the results show that the DV-HOP-RSSI algorithm not only has small positioning error, but also has good positioning effect and certain anti-interference, compared with the traditional DV-HOP algorithm and the improved DV-HOP algorithm.

The Super Low Frequency (SLF, 3~300 Hz) Channel Noise (CN) impulses are usually smeared by the transient effects in the receivers’ front-end stages, which will make the Blanking Nonlinearity (BNL) lose effectiveness. To solve this problem, based on the analysis of the waveform characteristics of transient response of impulses and the mechanism of performance reduction of the BNL, in conjunction with the consideration that the Local Variance Domain Transforming (LVDT) is capable to intensify the impulsiveness, an adaptive BNL based on LVDT is proposed. The Detection Threshold (DT) for the CN impulses is formulated based on the Constant False Alarm Rate (CFAR) principle and then the DT optimizing principle for the BNL is given. Simulations and real tests show that the proposed method outperforms other common NLs in terms of SLF CN impulse suppression. Since such method needs not to assume the CN model or estimate its parameter, thus is a blind suppression method, therefore, it is more practical.

To solve the problem of poor identification performance of coherent signals in the presences of non-uniform noise and mutual coupling, a Direction of Arrival (DOA) estimation method based on non-uniform noise covariance matrix and mutual coupling coefficient reconstruction is proposed in this paper. Using the Least Squares (LS) theory, the noise-free covariance matrix in the sense of mutual coupling can be recovered by iterative optimization approach. Then, the uncorrelated signals angle determined by applying the principle of signal subspace can be used to reconstruct the mutual coupling matrix, so that the noise-free signal covariance matrix after compensating the mutual coupling can be obtained. Finally, the spatial smoothing techniques can be employed to decorrelate the coherent signals, which enable us to apply the MUSIC algorithm to achieve DOA estimation. Simulation results demonstrate that, compared to the traditional MUSIC and DOA estimation algorithms, which only considers coherent signals, non-uniform noise or mutual coupling, the proposed method can suppress non-uniform noise significantly, overcome the decorrelation failure of the conventional spatial smoothing techniques in the case of mutual coupling, and improve the DOA estimation performance of coherent signals under the coexistence of both non-uniform noise and mutual coupling considerably.

To solve the problem of sparse expression of facial expression features, a facial expression recognition method based on separability assessment of the Discrete Shearlet Transform (DST) is proposed. DST is a relatively new image multiscale geometric analysis method. First, the DST transform is applied to the preprocessed facial expression images, and the transformation coefficients are obtained. Then, according to the separability evaluation function, the separability index in each direction and the scale coefficients are evaluated, and the low and high-frequency coefficients are fused on the best separability direction and scale. Finally, Support Vector Machine (SVM) is introduced to classify the facial expression. The experimental results show that the proposed method can select the best separability scale and direction coefficient as the feature, abandon the useless information, reduce the feature dimension and computation cost, therefore, the system is more efficient.

Super-resolution reconstruction algorithms based on convolutional neural network are confront the problems of small receptive field, disappearing of gradient information and slow convergence of the network. In order to solve these problems, an image super-resolution reconstruction algorithm based on residual connection convolutional neural network is proposed. Processing images in high resolution space results in the increase in network complexity. So, images are processed in low-resolution space to reduce complexity. The convolutional neural network is improved by introducing local and global residuals. Local residual promotes the flow of information and avoids the problem of gradient disappearing during network training. Due to the local residual connection, image information can be transferred to deeper layers. The global residual makes the network learn the residual information in images only, reducing network redundancy and accelerating network convergence. The receptive field is expanded by increasing the network depth, which makes the network learn more reconstruction information. The PSNR and SSIM of the proposed algorithm are comparable to other algorithms.

In the case of active interference, while the interference is suppressed, adaptive signal processing affects the sum beam and difference beam weight vectors, the sum beam and difference beam distortion occur, which make sum-difference angle measuring method produce a larger deviation, leading to the performance degradation of sum-difference angle measuring method. To solve these problems, a method of adaptive difference beam weighting for uniform subarrays is proposed, which could get high angle measurement performance while restraining interference. So the adaptive angle detection curve coincides with idea sum and difference angle curve. The proposed method is robust and can achieve good angle measurement under the presence of pointing errors and small snapshots. Comparison to monopulse angle measurement method for generalized sidelobe cancellation at subarray level reflects the advantages of proposed method. The computer simulation proves the efficiency of the algorithm.