According to the commonly used braking energy recovery rate and driving range increase rate of the braking energy recovery evaluation index, the vehicle specific system scheme and the braking force distribution are analyzed. Then, a braking energy recovery evaluation method based on wheel cylinder pressure is put forward considering the opening degree of the vehicle Controller Area Network (CAN) protocol. The wheel cylinder pressure signal is acquired by pressure sensor. Vehicle test is carried out and the results show that, with relative error rate less than 6%, the proposed method is feasible, which enriches the braking energy recovery evaluation theory.

The precise design of the dynamic stiffness of air spring is one of the key issues to accurately match the dynamic stiffness of the vehicle in the predevelopment stage. Considering the transition of the piston circular area, a method is proposed to get the accurate dynamic characteristics model of the air spring, which is based on the design of the piston shape in different stages according to the surface integral. The influence of the main design parameters on the dynamic characteristics of the air spring is studied. The experimental design sensitivity analysis and parameter optimization are carried out. Results show that the model can accurately predict the dynamic characteristics of the air spring and provide an economical and practical prediction method of the stiffness design of the air spring used in automobiles. When the air spring is in normal working pressure and recommended working height range, the optimization of structural parameters of the air spring piston can simultaneously take into account the comfort of vehicle itself and the comprehensive performance of the air spring and has certain economic value.

The traditional vehicle braking system is difficult to control the tire slip rate directly. In order to study the tire slip control mechanism of the Integrated-Electro-Hydraulic Braking (IEHB) system and the improve active safety performance of the vehicle, the physical simulation model of IEHB actuator and dynamic model of 7 DOF vehicle were established. Combining with the hierarchical control structure and taking the control method of dual loop, which constitutes slip rate and braking torque, the tire slip nonlinear controller based on IEHB is designed. Simulation test is conducted via co-simulation platform of MATLAB/Simulink and AMESim under the scenarios of active emergency braking process on high adhesion coefficient road and low adhesion coefficient road. Simulation results verify the proposed control method.

In this paper the vibration and noise of an electric bus were studied under the conditions of uniform speed and acceleration. The LMS instruments were applied to collect the vibration and noise signals of the electric bus. The source and the characteristics of the noise and vibration were studied by spectrum and order tracking analysis. The results show that the vibration inside the bus is mainly induced by the road excitation via the suspensions under uniform speed condition, while the motor noise is sharp under accelerating condition especially in the switch frequency. This study may provide useful information for reducing the noise and vibration of electric vehicles.

To recover the braking energy as much as possible without affecting the driving intention, an ideal braking force distribution strategy based on threshold value is proposed. First, considering the comprehensive optimal efficiency of the motor, battery and Continuously Variable Transmission (CVT), the CVT ratio control strategy is developed. Then, the relationship between the inertia torque and intention of driving brake is obtained based on dynamic characteristic analysis of the transmission system during braking process. Third, with the goal to increase the regenerative braking power, the target of CVT speed ratio and braking force distribution is modified using discrete exhaustive optimization algorithm in combining with the comprehensive efficiency of the optimal CVT ratio control strategy. Finally, a CVT optimal control strategy is proposed based on speed ratio limitation and motor coordinated control. Simulation results show that under different braking conditions, the regenerative braking control strategy enables the vehicle to follow the driver's intention effectively during braking, and increases the recovery ratio of braking energy and improves the driving performance.

To ensure better operation temperature of the battery pack for electric vehicle, a battery cooling system with heat pump was proposed, and the model and the system were built on MATLAB using theoretical and experimental description method. The numerical simulation results show that the proposed system can effectively meet the battery cooling requirement at high temperature and high working load, and the heat pump has obvious auxiliary cooling effect. If the battery radiator is placed in front of the condenser the cooling effect is better than that the battery radiator is placed behind the condenser. In addition, the impact of the liquid flow rate is more obvious than the rotational speed of the fan in the dynamic cooling process. Therefore, the pump should serve as the main regulation factor in the battery cooling process, especially under high temperature and high load conditions.

Hepatic trauma induced in traffic accidents is one of the most common injury patterns, causing high fatality rate. Human liver biomechanical responses associated with frontal impacts, oblique impacts were studied using simplified Chinese human body Finite Element Model (FEM), assembling with more geometrical-accurate liver model and an average Chinese adult male model from high resolution CT data. The model was than applied for studying liver dynamic response and injuries in frontal and oblique pendulum impact simulations at low, middle and high velocities. Analysis of dynamic response and pressure distribution with high injury risk on liver was performed. Pressure and peak strain that may induce hepatic injuries were computed from the model simulations and were analyzed about the correlation with global parameters, such as thoracic deflection, viscous criterion value, contact force. It indicates that using single variable representing liver response is questionable under different loading boundaries.

In order to overcome the difficulties caused by the difference of road surface conditions to the braking system, and improve the control of pressure hysteresis, a road peak adhesion coefficient identification algorithm was proposed and a single neuron PID pressure controller was designed. The influence of the road surface on the braking performance was deeply studied. The maximum braking force based on peak adhesion coefficient and the maximum strength brake of ABS were compared. The effectiveness of the identification algorithm was verified by Matlab/Simulink simulation, and the two kinds of pressure controller were compared. The performances of the two braking systems under different road adhesion conditions were analyzed. Results show that the identification algorithm has higher precision that the error is within 5%. The designed controller has faster response, smaller overshoot and no static error. With the deterioration of road adhesion condition, the ABS braking performance is decreased, while the braking system based on road identification still has good braking effect, which shows obvious advantages.

The kinematics model and solving algorithm of turning trajectory of B-double vehicles are proposed. The trajectories of the front and rear axles of the tractor, the rear axle of the first semitrailer and the rear axle of the second semitrailer are simulated using MATLAB. By changing the structurla parameters including the wheel base of the semitrailers and the position of the articulated point, comparative analysis of the trajectory of each axle center is carried out. Results show that by control of the corner of rear axle of the semitrailer, the handling stability and the flexible steering of B-double vehicles can be improved.

Bus hook turn method can guarantee the left turn operation of buses at the left-turn prohibition intersection. It is meaningful to improve the intersection capacity and maintain the operation efficiency of the bus system. To solve the problem of the queue bus spillovers at waiting areas and the waste of all-red clearance time at such intersection, a layout scheme of three kinds of detectors is proposed based on the actuated control logic. The queue spillover automatic detection method, the automatic determination method for all-red time and the adaptive signal control flow are developed. An intersection at Shenyang City is taken as an example and filed data are collected. Simulations and evaluations are conducted in VISSIM software. Results show that in peak time the proposed method can reduce the average vehicle delay by 10.8% and reduce the maximum queue length by 9.8%. In off-peak hour, the average vehicle delay and maximum queue length can be reduced by 13.1% and 17.2% respectively.

A Hybrid Dynamic Lane Operation (HDLO) is proposed for signalized intersection along an urban arterial to make best use of available arterial space and efficiently serve periodic unbalanced directional traffic. In accordance with the predominant direction of travel that varies over time, reversible lanes, changeable approach lanes, and conventional lanes are placed in combination on the opposing approaches along the arterial to channelize intersection in a traditional way. The movement of traffic is managed by traffic control device and signal timing techniques that are commonly used. Directions of travel on reversible lanes and changeable approach lanes are reoriented by traffic event-driven cooperation of road zipper and changeable lane-use sign. Simulation of peak periods is carried out and results indicate that, in contrast to existing lane operations, the proposed HDLO can improve intersection performance and balance the use of bidirectional arterial space. The operation advantage of HDLO becomes more significant when the intersection is operated near or over the capacity.

In transient dynamic congestion, there is a strong interaction among the intersection pedestrian. The conventional mechanical evacuation method lacks friction constraint parameters with changeable instantaneous dynamic characteristics, and assumes that there is no contact among the pedestrian. This results in poor evacuation process efficiency, poor stability and poor distribution equilibrium degree. To solve the above problem, a pedestrian transient dynamic congestion evacuation model based on cellular automata is proposed. First, the mechanical constraint parameters of the transient dynamic congestion are obtained. Then, under the constraint of these mechanical parameters, the traffic flow of the longer evacuation path is transferred to the shortest path. Finally, the cellular automata is used solve the transient dynamic optimal state of the user. Simulation results show that, applying the proposed model to multiple evacuation ports and single evacuation ports, the pedestrian congestion evacuation efficiency, pedestrian compliance rate and the equilibrium degree of evacuation distribution are improved.

This paper presents an algorithm for short-time passenger flow prediction in urban rail transit based on Moving Average (MA) method. First, the moving parameters of MA are determined, then, the direct forecasting passenger flow time is obtained by averaging the historical data. Second, the predicted time series of passenger flow are obtained by modifying the direct forecasting results with the actual-time passenger flow data. Experimental results suggest that the MA method is superior to Support Vector Machine (SVM), Back Propagation Neural Network (BPNN), Wavelet Neural Network (WNN) and the combination of Wavelet and SVM (WS) both in prediction accuracy and computing speed. Therefore, the proposed algorithm can be applied to short-time prediction of complex passenger flow time series.

A new method of directed weighted network is proposed to solve the problems of assembly line design and assembly line balancing. The structure of the product (car model) is analyzed on the existing linear assembly line in the laboratory, the assembly line is designed. The process time and preceded relationship between the tasks are obtained, and the line balance is not high. In order to improve the production efficiency, the mathematical model of the product assembly process is established. The weight of the priority relationship in the model is calculated using directed weighted network theory. The computer program is calculated and the assembly process of four wheel drive model is reconstructed by the heuristic algorithm. The U-shaped assembly line is compared with the existing linear assembly line, which shows that the balance rate is increased by 9.4%, thus verifying the superiority of the model and the algorithm. The weighted network theory improves the operation efficiency and provides a new method for calculation of the balance problem of large assembly line.

In order to improve both the high temperature stability and low temperature crack resistance of asphalt at the same time, nano materials and basalt fiber were used to modify the asphalt. the high temperature stability and low temperature crack resistance of the composite modified asphalt were evaluated by cone penetration test, softening point test, ductility test, viscosity test, dynamic shear rheological test and low temperature beam bending creep test. Results show that the incorporation of nano materials and basalt fibers can effectively improve the shear strength and cohesive force of asphalt and enhance the high temperature resistance to permanent deformation. The incorporation of nano materials and basalt fibers reduces the low-temperature ductility while increases the low temperature creep rate of asphalt. The low temperature stress relaxation ability of asphalt is also enhanced and low temperature performance is improved. The incorporation of two kinds of modified materials has a superposition effect on the improvement of the asphalt performance, and the application in construction of asphalt pavement is beneficial to further improve its service life.

For the three-span rigid-frame continuous bridge with main span of 160 m in Changbai Mountain International Tourism Resort, the dynamic characteristics in each cantilever construction stage was calculated using Midas Civil software, and the characteristics of the vibration mode and frequency of each construction stage were analyzed. First, according to the climate condition at the bridge site, the gust wind speed was calculated, then the wind field area was simulate using Midas FEA software CFD module. Second, the three component coefficient of the maximum cantilever state of the bridge structure was calculated, and then the static gust was obtained. Finally, the simplified calculation method of buffeting force is adopted to analyze the buffeting force of the bridge under the maximum cantilever state, and the safety of the structure under the maximum cantilever state and wind load was evaluated.

In order to control the cylinder pressure of free-piston engine generator, a variable-load control method is proposed. The control of the total impedance of the system circuit is achieved by the construction of a variable equivalent load, so that the loop current is changed, and then the electromagnetic resistance acting on the mover changes. In order to make the electromagnetic force change in accordance with the expected law, a closed-loop control strategy is designed. Simulation results show that, using this control strategy, the proposed variable-load control method can stabilize the piston motion, thus achieving a stable cylinder pressure.

To develop a piezoelectric common rail injection, the piezoelectric stack actuator was selected as the driving element. In order to exploit the potential of this kind actuation, the thermo-electro-mechanical behavior of the piezoelectric actuator was investigated by experiment, with maximum voltage of 150 V under different mechanical loads and different temperatures. Results show that, as the load increases from 200 N to 2000 N, the displacement increases first, and then decreases. In the temperature range of 25?C to 80 ?C, the displacement increase with the temperature, but this trend gradually slows down. A mathematical model based on the hysteresis and temperature effects is applied to simulate the experiment data, and the simulation results show reasonable agreement with the experiment results.

The research focuses on the determination of reasonable pre-tension of crawler traveling gear of hydraulic excavator. The virtual prototype model of the excavator was established in the multi-body dynamics software RecurDyn. The pre-tension curves of the crawler traveling gear were obtained by experiments under three working conditions, horizontal straight, pivot steering and differential steering. Experiment results verified the accuracy of the virtual prototype model by comparing with the simulation results. By simulation analysis of three typical working conditions, horizontal straight, climbing, and pivot steering, the changes of the track tension during movement on wheel were obtained with the transformations of pre-tension of the crawler driving device. According to the judgment of the fluctuations based on the standard deviation of the tension curves, the reasonable proportional coefficient between the pre-tension and vehicle weight is received.

In order to study the effect of abrasive grain size on abrasive flow polishing, common rail pipe was selected as the research object. By calculation the state of abrasive grain flow under different grain sizes, the discrete phase analysis of the flow field was carried out from the dynamic pressure, the velocity field, the abrasive grain trajectory and the turbulent kinetic energy. From the numerical analysis shows that, with the increase in the grain size, the dynamic pressure, the velocity and the turbulent kinetic energy are reduced and the grinding effect is weakened. The trajectory of the abrasive grain can be used to predict and theoretically guide the optimal control of the abrasive grinding path and the production process, thus achieving effective and accurate polishing. The surface roughness and the surface topography of the common rail pipe were measured before and after abrasive flow grinding. The surface roughness was 3.401 μm before the grinding and was 1.138 μm after the grinding. This verifies the effectiveness of abrasive flow polishing workpiece with inner channel structure, and confirms the correctness of the numerical analysis. This work may provide theoretical support for the development of abrasive grinding technology.

In order to meet the requirement of commonality and effectiveness in solving the intensity factors of piezoelectric bimaterials with interface crack in practical engineering, based on piezoelectric interface facture mechanics, an explicit extrapolating formula of intensity factors for in-plane and anti-plane interface crack was derived. The interface crack-tip displacement field and electric filed of the piezoelectric bimaterials were simulated using the electromechanical coupling computation methods. Substituting the crack displacement and the electric potential jump at the back of the crack tip into the explicit extrapolating formula, the intensity factors of piezoelectric bimaterials with interface crack were solved. Considering a center interface crack in double piezoelectric plates, the intensity factors under different loadings, different numbers of elements, and different mesh refinement methods were discussed and compared with the analytical solution. Results of numerical examples show that the proposed method has the advantages of simple calculation and high accuracy.

In order to build the analytic model of soybean kernel using Discrete Element Method (DEM), the shape and size of soybean kernel are measured and analyzed. It is found that the shape of a soybean kernel can be approximated to an ellipsoid. The three axis dimensions follow normal distribution, and there exists a certain correlation between the primary dimension and the secondary dimension. On this basis, a soybean kernel modeling method is proposed based on the sphere filling. The shape of the soybean kernel is approximated to an ellipsoid, and the three axis dimensions are determined as the mean values of measured soybean kernels. The primary dimension of the analytical model of a kernel, which follows normal distribution, is randomly generated, and the secondary dimension of the kernel is calculated according to the relationship between the primary and secondary dimensions. As a result, the assemble model of the soybean kernels with different size and distribution is built. Two varieties of soybean kernels are selected as the research objects. By comparison between the simulated results and the experimental results of the packing test and the self-flowing screening test proposed in this work, it is shown that, when the amount of filling-spheres is five, the simulation results are in good agreement with the experiment results, thus, preliminarily verifying the effectiveness and feasibility of proposed modeling method.

To improve the autonomy of suspension insulator cleaning robot, a behavior planning method based on hierarchical structure and finite state machine is proposed. According to the structural features of the robot, the environmental characteristics and the cleaning task, the behaviors of the robot are classified into behaviors and combinational behaviors. The methods to describe these behaviors are presented and the finite state machine is used to manage the sequences of the combinational behaviors. The exceptional handling mechanism is introduced to ensure the reliable operation of the robot. Simulation and laboratory experimental results verify the correctness and effectiveness of the proposed method, which can be applied for behavior planning of the robot to perform suspension insulator cleaning task.

A heuristic algorithm A^* for measuring the similarity of multiple graphs is proposed. In this algorithm, the multiple graphs are represented as multiple sequences, then the number of all common subsequences in the matches of the multiple graphs is calculated, and this number is used to measure the similarity of the multiple graphs. This algorithm avoids the redundant calculations in non-matches of multiple graphs, maximizes the heuristic function value of all common subsequences of the suffixes sequences, limits the search nodes to the subset of matches between two sequences, thus reducing the number of the calculation nodes. Compared with the existing graph similarity methods, the proposed algorithm can not only measure the similarity of multiple graphs, but also simplify the calculation process. The similarity of multiple graphs can be quickly measured using this algorithm with the boot of the heuristic information.

The main challenges in a single target persistent tracking are the factors such as the change of the target pose, similar background and occlusion, which account for the difficulty in solving the drift problem. Based on Tracking-Learning-Detection (TLD), we propose an improved Li Tracker Using Accelerated Proximal Gradient Approach (L1APG) tracking algorithm. First, we add occlusion detection to the L1APG tracker. Then, we transform the occlusion problem into the weight of the target template and the background template coefficients. Finally, the original tracker is replaced by the improved L1APG tracker in the traditional TLD algorithm, the coefficient weights vary adaptively according to the degree of occlusion in real time, which contributes to improving the tracking results effectively. Experiments show that, compared with TLD and L1 tracking algorithms, the proposed algorithm can better deal with the occlusion and drift problems and possesses better stability and robustness.

In this paper a quantum k-means algorithm is proposed by integrating the quantum paradigm to improve the efficiency of traditional k-means algorithm. First, each vector and k cluster centers are prepared to be in quantum superposition, which are then utilized to compute the similarities in parallel. Second, the quantum amplitude estimation algorithm is applied to convert the similarities into quantum bit. Finally, from the quantum bit the most similar center of the vector is obtained using the quantum algorithm for determining the minimum. Theoretical analysis shows that, compared with the traditional quantum algorithm, the time complexity of the quantum k-means algorithm decreases under given condition and the space complexity diminishes exponentially.

An evolution model of software networks based on local events is proposed. The model uses objective-oriented software systems as research object and improves the BA scale-free network model. By adding local events, such as addition of nodes, addition of edges, removal of edges and rewriting edges, the proposed model simulates the evolution of software networks. Simulation results show that the degree distribution of the network generated by this model follows power law distribution. The model can simulate the evolution process of real software, and the degree distribution is in agreement with that of real software network, which validates the proposed model. Meanwhile, the proposed model can simulate and evaluate the structures of object-oriented software networks.

In order to handle the safety problems in the Identity-based Encryption (IBE) key schemes associated with the existing cloud services, an IBE scheme based on quantum secret keys and cloud service is proposed. First, the quantum encryption technology is introduced to handle the safety problems of information transmission between the private key generator and the cloud services, involving user privacy information and revocation lists, etc. Next, the quantum time token mechanism is proposed to safely distribute and manage users' secret keys. Finally, the user revocation management is safely and effectively achieved by sharing the quantum secret keys and using the cloud services. Comparing with the existing schemes, the proposed scheme can make full use of the advantages of the quantum encryption technology and the cloud computing services. The secret key management and user revocation of the IBE public key scheme can be more safely and effectively done, and multiple safety hazards can be avoided with better safety and practicality in the cloud computing environment. Experimental results in the real quantum key distribution and the cloud computing environment demonstrate the feasibility and effectiveness of the proposed scheme.

The correlation indicator, Maximum Information Coefficient (MIF), which can pervasively measure the nonlinear relationship, is introduced to solve the problem of inaccurate selection of the near-neighbor set of the target users. The indicator is employed to measure the similarity between users. First, the near-neighbor set target users is selected based on a given threshold. Then, the personalized SVM prediction model is built with the attained near-neighbor set as the training set to carried out scoring prediction for the interesting items of the target users. Simulation results show that the near-neighbor set selected by the MIF Measuring is more accurate than that selected by the Pearson Measuring, and has the merit of insensitive to the threshold.

In order to solve the problem of Android malware spreading, this paper proposes an Android malware static detection model. First, the attribute value of three label items is selected in AndroidManifest.xml file as the feature attribute. Second, the Information Gain algorithm (IG) is used to optimize the feature attribute, then a corresponding feature vector set is generated according to the optimization results. Finally, four kinds of machine learning classification algorithms in Waikato Environment for Knowledge Analysis (WEKA) are used to detect and classify the feature vector set. Experiment results demonstrate that the proposed static detection method has better detection and classification performance.

During data transmission in Wireless Sensor Network (WSN), the nodes are always faced with the problems of energy depletion and data congestion. To solve these problems, this paper establishes a mathematical model of node reliability, and the optimal value of the node degree is obtained by maximizing the node reliability and the network lifetime. Then, in accordance with the optimal value of the node degree, a Topology Control based on Node Reliability (TCNR) algorithm is put forward. Simulation results show that the TCNR algorithm can greatly reduce the congestion degree in data transmission, enhance the robustness of the topological structure, and effectively prolong the WSN lifetime.

To solve the problem in multiple workshop manufacturing that there are single complex products with multi-procedures ended at the same time, a reversal sequence integrated scheduling algorithm is proposed. First, a reverse batch scheduling strategy is designed to solve the problem that migrate virtual processes for positive sequence scheduling can lead to equipment resource idle operating complex. Then, in order to reduce the process migration and keep the balance of load among the workshops, the workshop identified strategy is proposed to assign all the processes to the workshops. Finally, in order to satisfy the special constraints of multi-procedures ended at the same time, a strategy of reversal sequence starting together is designed to determine the reverse starting processing time of each virtual procedure group. A case study verifies that the proposed algorithm meets the special constraints, and the completion time is shortened with less process migrations.

To combine the advantages of line stocking and kitting, a novel line-integrated supermarket model was proposed to further improve the part feeding process in mixed-model assembly lines, and the task assignment of logistic workers and part feeding period were optimized. First, the interdependent task assignment and cyclic part feeding problems were described. The mathematical models were built to minimize the total cost of logistic workers hiring and parts delivering. Then, based on the theorems, the dynamic programming embedded with heuristics was adopted to obtain the optimum for small scale problems. For medium and large-scale problems, a Modified Harmony Search Algorithm (MHSA) was constructed to generate satisfactory solutions. To accelerate the searching process, the harmony memory was repeatedly divided and regrouped. A local search method, crossover and mutation were employed to explore the searching space, thus to overcome the deficiencies of the original algorithm, such as limited search depth and the tendency to trap into local optimum. Finally, simulation of the modified algorithm was carried out and the results were compared with that of other evolution algorithms. The rapid operation speed and effectiveness of the proposed method were verified.

Data dynamics can not be ignored in the design of recommendation algorithm. Most of the traditional static text modeling methods are based on the basic assumption of exchangeability, but the dependence of the data in the covariate space is neglected. To solve this problem, a new dynamic function DPM recommendation model is proposed based on process model. This model improves the traditional Diky mixture model in dynamic data modeling. The space relationship of the parameters and covariates of the dependent Dirichlet process is created, while the Dirichlet process still belongs to the marginal distribution. Application of functional Dirichlet process can carry out effective modeling for mixed model assembly for disappeared and the change of the parameters, and can be used as a dynamic prior into parametric mixture model. Simulation results show that, compared with traditional de Lickley process, the proposed algorithm has more obvious advantages in doing a priori topic model.

For millimeter wave communication with uniform linear Antenna Arrays (AA), a 3D Multiple Input Multiple Output (MIMO) channel model is established to study the relationships between the capacity and the antenna separation, the carrier frequency, the antenna number an the distance between the transmit and the receive AA. First, utilizing the channel orthogonality, the optima antenna separation corresponding to the achievable maximum capacity is derived. Then, the performance variations caused by the increasing antenna number and the deviation of the antenna separation are analyzed respectively. Simulation results show that, when the distance between the transmit and the receive AA is within a certain range, the millimeter wave communication system can achieve the optimal capacity, which exceeds the capacity of the conventional MIMO system in Rayleigh fading channel.

Long due time in traditional Tong detection algorithm results in slow speed of detection and confirmation. In order to solve this problem, an improved Tong detection algorithm with double thresholds is proposed. By comparing the detection statistic with two-stage thresholds, the detection condition of the algorithm can be met more quickly after a few logical operations on the counter, which greatly reduces the detection time. The false alarm probability, the detection probability and the average dwell times of the proposed algorithm are analyzed and verified. Simulation results indicate that the proposed algorithm can maintain the detection performance and reduce the average dwell rimes effectively if a second threshold is set suitably.

To estimate the parameter of α-stable distribution signal, this paper proposes an algorithm based on the plus property and p-norm property of α-stable distribution. The characteristic exponent α is estimated using the relationship between the p-norm of original variable and the new p-norm of variable, which is the combination of several variables with the same α-stable distribution. The definition and the properties of the α-stable distribution are introduced, and then the principle and execution steps of the proposed algorithm are explained. The unbiaseness and uniformity of the algorithm in estimation of the characteristic exponent are discussed. Simulation results show that this method can estimate the characteristics exponent of α-stable distribution in a wide range without any more information of the distribution. It works well even Gaussian noise is added to the α-stable variable. The estimation results can support the subsequent signal processing algorithm with α-stable noise.

In order to solve the problem of over-subdivision of Fast Adaptive Segmentation (FAS) algorithm, a dynamic adaptive feature subdivision method is proposed. First, the Feature Processing Unit (FPU) is constructed to generate the dynamic subdivision factor for the irregular block. The subdividing depth of the feature block is calculated and the irregular block of the feature region is refined according to the Catmull-Clark subdivision rule. Then, the number of blocks is controlled for Graphics Processing Unit (GPU), the block buffer and subdivision table are established to deal with each level of the same type of blocks in parallel, which improves the speed of subdivision and rendering. Finally, the data structure of the traditional FAS is expanded to generate new subdivision tables and drawing tables, which support the dynamic subdivision and real-time rendering of blocks. Experiment results show that the structures of the improved subdivision table and rendering table can guarantee the dynamic feature of the subdivision and the real-time rendering. Compared with the traditional FAS, the proposed method can ensure the drawing accuracy of 3D surface model, and increase the drawing speed by 28%, which is better in real-time performance than traditional FAS.

Inspired sand scorpions in locating their preys, a novel bionic localization method is proposed to solve the problem of vibration event localization. Different from commonly used methods, a spike firing model is introduced to solve the problem of arrival picking. Based on the 3/1 configuration, each spike firing model fires different number of spikes. The vibration signal is coded into spike signal for orientation of the target using population vector code. According to the actual situation and the selectivity of the biological evolution, the distribution of the receivers is modified and compared with that of before modification. Also the bionic method is compared with the TDOA method in simulation environment. The orientation errors of the bionic method are smaller than that of the TDOA method. In addition, location experiment is carried out in an open area. The average error of landing orientation is 4.9310° for 76 sets of data, satisfying the orientation requirement.

In this research we designed a prototype of Sleep Apnea Syndrome (SAS) detection device based on the surrounded type pulse oximeter probe. This self-designed device was compared with clinical commercial finger clip type oxygen probe. Furthermore, SAS was preliminary detected and analyzed simulated apnea. The results show that the peripheral oxygen saturation (SpO2) measurement error of this self-designed device is less than ±1%, which meets the requirements of sleep monitoring, thus, can provide reliable data sources for SAS automatic detection model. This device provides a new choice fro SAS screening.