The occupancy data of a small region contains 29 loops in Beijing city was analyzed using generative topographic mapping through time(GTM-TT). The original data was mapped from 29 dimensional space to 2 dimensional space which is easy for visualization and clustered into different classes naturally. The statistics of states transfer show that the most probably transfer distance was within 1 step. At last the 89 days was clustered into 4 types using hierarchical clustering method according to the Euclidean distance between the tracks of different days in the latent space. It was found that the day of different type had different traffic process.

In order to acquire urban traffic information timely and accurately,a method of collecting trafficinformation based on inter vehicle communication was putforward.All the vehicles which can transmitand receive data by wireless communication were regarded as moving sensors on traffic network.Road side station can get traffic status by analyzing data collected by equipped vehicles.Moreover,aproto type system of WirelessTraffic Information Service System(WTISS)was designed,and WTISS simulation system based on Paramics,NS2,MatlabandMy SQL was constructed.The method of collecting traffic information mentioned above was verified on this simulation system.

Coordinated control regions were generated statically in the past, which can't accommodate to the changing traffic condition over the road networks. Thus, a new method is proposed to solve this problem by subdividing the road network into control regions dynamically using a hypergraphbased model. First, the platoon dispersion model is analyzed; a new concept called "coordination desire" is indicated to describe the requirement of coordinate control between adjacent intersections. Then, we designed a hypergraphbased model to express the road network. A widely used hypergraph partitioning algorithm HMETIS is employed to subdivide the hypergraph into partitions which correspond to the coordinated control regions. Finally, a case study based on the actual road network and practical traffic data is presented to demonstrate the effectiveness of our model.

To reduce traffic delays, relieve congestion, and improve the operation of urban traffic network, a phase sequence optimization model was presented based on the idea of maximization of twoway green wave bandwidth. Under 15 difference groups of scenarios, VISSIM was used for testifying the effectiveness of the optimization model. The simulation results show that the optimized scheme reduces the intersection delay by 2% ~ 29%, the number of stops by 14% ~ 28% and travel time by 1% ~ 9% , which greatly improves the operation condition of urban road network.

An international comparative analysis between Beijing and New York City is conducted, such as vehicle ownership, diurnal distribution, trip distance, trip purpose and distribution of persons per vehicle of private car. Based on difference of the two cities, the discrete choice model of factors influencing car usage and calibrate the parameters is developed with 2005 Beijing Household Travel Survey data. The estimation results and elasticity effects indicate that such factors as number of household bikes, working or not, monthly ticket, trip purpose, parking fee and trip time have negative effects on private car usage, while number of household cars, child, driving license, main user or not, company and trip distance make positive effects.

Based on user equilibrium, the complex traffic behaviors in regular networks, random networks, smallworld networks and scalefree networks were investigated. Within user equilibrium, BPR function was used as link cost function, and two types of link's practical capacity which is a parameter in the BPR function were analyzed in detail. In order to describe the congestion of the network, the relationship between the rate of links whose load is larger than its practical capacity and traffic demand was studied. Finally, the load distribution in the above four urban traffic networks were discussed.

A waiting duration model of pedestrian crossing was proposed by using hazardbased duration models based on observed data. The framework of baseline hazard adopted parametric shape and nonparametric shape. The factors that could effect crossing behaviours were analyzed, particularly for crossing in violation of traffic rules. The results show that the nonparametric model can present an accurate description of waiting behaviours. The crossing behaviours have significant dependence on time, the hazard of traffic violation increase when time elapse. The results provide a useful reference for pedestrian crossing analysis and traffic behaviour simulation.

In this paper,link transmission model(LTM)was applied to simulate traffic flow under various traffic demands,by which we can investigate the influence of the length of channelization area on the performance of traffic flow.An approach to calculate flow density anywhere of a link was proposed.An ovelmethod was developed to calculate congestion delay,which is beneficial to evaluate the performance of traffic flow.Numerical examples show that the level and the uniformity of traffic demand are the key factors that impact the length setting of channelization area.Inaddition,the length setting of channelization area under signal control was discussed.

Based on factors of pedestrian and vehicle operation, the delay model on crossing location of pedestrianvehicle interference system was established according to crossing demands of pedestrian and vehicle. Several influence levels along with location of crossing were analyzed. Taking advantages of VISSISM software and one 2lanes branch road in Xicheng District, Beijing as an example, one best location of crossing was developed from the 7 choice locations. The results show that average delay of pedestrian is directly related to the walking distance from starting/ending points to crossing of pedestrian, but indirectly related to average delay of vehicle. The average delay of pedestrian is much bigger than the average delay of vehicle.

The traffic system with two lane mainroad and one lane on-ramp was simulated and analyzed by using the modified weighted probabilistic cellular automaton model.The conformity effect of the vehicles on the ramp was considered.The phase diagrams with and without the conformity effect were analyzed and compared to study the influence of conformity effect on traffic flow characteristics.The variations of the capacity at upstream the on-ramp,the on-ramp and down-stream the on-ramp with respect to the conformity probability were analyzed,when main road and on-ramp both are saturated.The results indicate that conformity effect can weaken the bottle necks effect caused by the main road on the on-ramp to some extentand increase capacity of on-ramp.

Based on the analysis of influencing factors of busstop routes capacity, a mathematical model describing relationships between bus parking time and traffic flows and speeds under different congestion conditions was established using different traffic flows on highspeed road stretches to express the congestion degree. Through statistical standardization of actual investigational data and parameters, busstop routes capacity under different traffic flows and speeds was calculated and, with a bus stop (3rd Ring Road NorthJimenqiao West Bus Stop) as an example, rationality and effectiveness of the model was validated.

In this paper, we propose a new function in which the summation of representative utility and random term is replaced by their multiplication, and derive a new NL model. Furthermore, letting the utility function be a mixture of summation and multiplication, namely the representative utility is decomposed into two added parts and the random term two multiplied parts, we derive the NL model and obain the marginal choice and conditional choice probabilities of an antenative. This study extends the NL mmodel applications greatly.

In this paper,considering different evolution behaviors in acceleration and deceleration of driving and the reinforcement learning,we propose a modified cellular automaton model in which the random acceleration behavior evolves with the driver's remembering to historic experience and the traffic condition around.Simulation results show that the proposed model can reveal the influence of driver’s reinforcement learning on traffic and generate more realistic traffic phenomena.

The efficiency losses of multiclass stochastic user equilibrium traffic assignment under road pricing against system optimization (SO) was investigated. The upper bounds of efficiency losses were analytically derived by variational inequality method. It is shown that whether using the timebased or monetarybased measures for decision making, the upper bound against SO always depends on the class of link travel time functions, the users' socioeconomic characteristics, the road pricing scheme, the network complexity, the total traffic demand and the degree of perception error.

A two route model which nears to real traffic systems  were studied by comparing three kinds of information feedback. The simulation results show that the optimal strategy depends on the traffic situation. Thus, the road manager should adopt reasonable feedback according to the real traffic situation, so that the foundational equipment can be fully used.

The effects of different traffic lights period on the traffic flow were studied based on two dimension BihamMiddletonLevine (BML) model. The simulation results in BML model show that, the average velocity of the vehicles has a sharp change from highspeed phase to lowspeed phase at the certain critical vehicle density with the increase of vehicle density, and this critical density do not strictly decreases with the increase of the traffic lights period. A meanfield theoretical analysis was made,and its results basically agree with the simulation results. Finally, the nonmonotonic phenomenon was futher discussed.

Traffic flow at a signal controlled T-shaped intersection where there are three input directions of vehicles was investigated by the cellular automata model. Because in the previous models some vehicles at intersection only have one steering way, a more realistic model which vehicles on each input lane have two steering directions was introduced. The three phase traffic signal with fixed period was used to control the conflicts of vehicles. The phase diagram and capacity of the system were investigated and compared with previous unsignalized models.

Given urban land use and road networks, computational experiments of evaluating road network traffic conditions were studied based on Artificial Transportation Systems (ATS). Components of ATS and the function of each component were introduced. An example was given to illustrate how to evaluate road network traffic conditions based on ATS. The percentage of congestion roads in the road network and the maximum number of vehicles existing in the road network during the rush hour were used to reflect road network traffic conditions. The computational experiment results can provide decision support to traffic management sectors.

A one dimensional totally asymmetric exclusion process with particles adsorption/desorption in the bulk was studied. At the same time.A global constraint on the total number of particles was consideried. In this model, the entry rate and attachment rate of particles into the lattice are influenced by the particle number available in the reservoir. This model is used to simulate a linear city transportation system with one main artery, which connect the residential area and CBD. The phase diagrams of model in different total particle numbers are shown. It is interesting to find that Phase coexistence can be observed and phase diagrams are affected by constant supply of particles. A meanfield approach is used to interpret the numerical results obtained by Monte Carlo simulations.

The effect of acceleration a andrandomization Dinare fined NagelSchreckenberg traffic flow model is investigated.It is found that when a≥D,there is no qualitative change of the fundamental diagram and spatiotem poral pattern.In contrast,when a<D,qualitative changes of traffic flow characteristics appear,and the congested traffic flow exhibits the characteristics of synchronized flow.Based on this finding,we study spatiotem poral pattern sinduced by an on-ramp,using the refined model with a<D and considering slowtostart effect.It is shown the patterns are consistent with the results of three phase traffic theory.Our workis expected to shed some lighton the mechanism of synchronized flow,which might be related to the value of a/D.

On the basis of reliability theory, how to delimit the public transportation terminal and the transfer reliability was discussed. Based on analysis of quondam transfer reliability models of the public transportation terminal,the transfer reliability model of public transportation terminal which include transportation multimodes and multiline transfer was established. The priority of the transit lines was determined by passenger numbers, and the weight percentage of transit lines in the public transportation terminal was calculated by the proposed method. Combined with the specific case, the calculation results was compared with the quondam model, proving the model applicability. Finally, some advice was put forward to improve the transfer reliability of public transportation terminal in line with analyzing the reliability model.

Because of the deficiency of conventional aggregate trip distribution model only considering single land use attraction variable, without including qualitative variables and hardly reflecting individual choice behavior, a new work destination choice model with aggregate alternatives, introducing size variables, is established in this paper. The application of this new model indicates that it contains multiple scale variables measuring land′s attraction degree as well as other discrete dummy variables. Furthermore, the model reflects individual′s destination choice behavior by introducing specific varibles of individual sociodemographics and mode chocies which interact with impedance.

Evaluating the congestion degree of road network comprehensively with traffic congestion degree was put forward. Using the mathematical theory of analytic hierarchy process and fuzzy comprehensive evaluation, and the results of userequilibrium traffic assignment with turn delays, the calculation method of traffic congestion degree aiming at the phase of traffic planning was also provided, based on the researching routing of pointlinearea. At last, according to the numerical example, this paper provided the concrete steps of calculating the traffic congestion degree and illuminated the feasibility and validity of evaluating the road network congestion degree with traffic congestion degree.

The structure model of system engineering modeling method will be applied to clarify the logical relationship between different factors. At same time, the mathematical relationship between various factors of the structure model will be established by fitting current data of traffic investigation. Furthermore, the mode split prediction method based on the structure model of various factors will be set up. The result of the case analysis with less errors illustrates this method is easily applicable and of great value.

The shareduse paths used by pedestrians, cycles and electric bicycles without physical separation or height difference was studied. Firstly, a basic cycle traffic conflict model was put forward to describe cycle passing and meeting events. Secondly, some special data collection methods were designed for the object of study, and five typical trails of slowmoving traffic were selected to perform the investigation in Nanjing city in China. Thirdly, electric bicyclescycle passing conflict model, cyclecycle passing conflict model, cyclepedestrian passing conflict model and cyclepedestrian meeting conflict model were set up respectively according to data collected. Finally, an intensity conflict model was built considering the effect of shareduse paths' width. The results of sensitivity analysis show that the value of all types of traffic intensity conflicts decreased with the increment of path's width, while cyclepedestrian passing and meeting intensity conflict contributed much more to the total impact than bicyclescycle passing intensity conflict and cyclecycle passing intensity conflict in shareduse paths.

Aimed at the case that there is no place for opening up bus lanes limited by the road conditions and traffic conditions in the center area of city,in order to carry out the signal control of urban public transit priority on the seroads,the traffic flow detectors are setted up the export of upper intersection,the time interval of bus and other big vehicles to go through the stopline of downstream intersection is estimated based on the realtime collection data,a series of bus operation schema at the stopline of downstream intersection are formed by merging each bus’time interval,by comparing with the signal operation schema of arterial direction to determine the green signal extension orredlighta head of arterial direction in order to achieveurban public transit priority.Traffic simulation results show that the mean delay of all vehicles reduce and the bus’delay decrease larger inaddition to the mean delay of crossroad vehicle increase.

This paper uses fuzzy support vector machines (FSVM) algorithm to classify urban road traffic status. Fuzzy membership handles linguistic expression bias of human perception and uncertainty in status parameter partition scale. Meanwhile, integrated SVM learning ability can solve the limitation that pure fuzzy classification method can′t train sample data. The one against one approach in FSVM is applied for multiclass classification. Finally, a microcosmic simulation work is performed for an experiment data example, and the results indicates that FSVM method can reduce influence of noise sample data and provide better classification accuracy.

Based on travelers via road,the trip distribution and traffic assignment model was developed for planned special events parking.Logit  based utility model was employed to present travelers’parking choice behavior,which taking into account driving time,parking fees,bus costs,bus time and walk time as main factors.By transferring link cost between parking and stadium to parking attraction,the singled estination network became a multi destination network.The equivalent mathematical programming model was developed to describe the equilibrium as signment,and its equivalence was proved.The experiment indicates the proposed model can describe road travelers'parking behavior.

This paper study stops at signalized intersection using macroscopic traffic model. Through the analysis the startwave and stop  wave in front of the stop line, it is concluded that the absolutely onestop condition don't exist. And the relationship between accumulatedischarge mode and stop ratio is researched. Trajectory of vehicles is classed according to the stable condition and the stop ratio is studied. The designed numerical example shows that stop ratio model based on Stable Theory is always over estimate the stop ratio, and the bigger the saturation is, the bigger the result is overestimated. The reason is that it adopts the point queue assumption.

This paper attempts to develop a kind of quantificational method to evaluate the level of service for the facility. The basic idea which was derived from westerner is used for reference. According to the specific traffic situation in China, the situaiton which a bicycle is totally hindered by a group of sidebyside pedestrians is highlighted in this paper. The probalistic models for the calculation of hindrance frequency are developed for both the same and opposite direction. The evolution trends of hindrance frequency according to the pedestian flow and the width of sidewalk are illustrated by the given example. The research results can be applied to provide theory guide to the planning and designing of shareduse sidewalks.

The operation processes of vehicles are analysed for both direct leftturn and rightturn combined u-turn(RTUT) at the edian opening at unsignalized T intersection. The calculaiton models for travel time and stop delay are developed to describe these two situation, which are based on the gap acceptance theory. Both the status of tranfic flow of major road and the geometry parameters of the section are involed in the models, which is different from other research findings. Hence, the models could be adopted for the evaluation for all kinds of implements of RTUT. Moreove, the comparion results which are derived from the calculation example are consistent with the known findings.

Based on fixed green time determination method and actuated control principle, this paper designs control process of phase splits, establishes the maximum elastic green calculation method and green real time optimization method for actuated coordinated control. The example results show that the method has good practicability.

A novel method is proposed to study representation and recognition of traffic event based on context in this paper. Firstly, foreground moving objects and their motion parameters are extracted by video detection and tracking; and then context is divided into three classes: spatial context, temporal context and object context by the surroundings traffic event occurred in. From these work, the concept of event cell is developed by the integration of moving objects properties and contextual information. Basic event and complex event are represented according to the execution scheduling of event cells. The event recognition method is presented at last. Based on video surveillance of real scenes, the method is validated that it can be well applied to recognize traffic event automated, and it could become an efficient technique support for the management and control of intelligent transportation.

For the defects of traditional artificial recognition and record method of traffic conflict, traffic conflict automatic discrimination at nonsignalized intersection based on video processing is presented in this paper. Firstly, foreground moving objects in the image and their spatial and temporal trajectories are extracted by motion detection and motion tracking; and then, the motion behaviors are obtained based on objects types and moving data. Traffic conflict can be discriminated in terms of two traffic conflict discrimination rules which are concluded combined with classic traffic conflict definition. Based on video processing captured from nonsignalized intersection, the method is validated that it can be well applied in realworld.

Based on the lucubration of info demand in the Urban Traffic Flow Guidance Systems, A new technology applicated to the acquisition of travel time was presented. Firstly, calculates the mean travel time by means of inductive coil detection and probe vehicle respectively. Then carry out the shortperiod peidiction thanks to the Adaptive Weight Exponential  Smoothing Ultimately empolders a frequent fusion arithmetic of travel time under a series of reliability degrees based on both inductive coil detection and probe vehicle. The experimental results confirm the performance of the technology as it can offer the travel time exactly and efficiently used as the input datas of he Urban Traffic Flow Guidance Systems to satisfy the info demand of travellers.

For a better solution of the congestion shifting problem in DDVGS Distributed Dynamic Route Guidance System),this paper develops a constrained K-shortest paths algorithm within a dynamic restricted searching area inconsideration of a real city roadnets patial distribution features which is suitable for the travelers and realizes its program on the basis of MAPX Control.The actual effect of this algorith mistested with the microsimulation tool VISSIM.Some conclusionshavebeendrawn as follows,the constrained K-shortest path salgorithm can not only decrease the searching scale and improve its running efficiency but also efficiently balance the traffic flow and prevent the congestion shifting problem,so that both the travelers and the whole system could benefit a lot from this.

Aiming at the problem that the turning travel time could not be collected based on the traditional link concept with an illegible border, a new link segmenting method was proposed to improve the performance of the travel time collection system based on GPS/GIS probe vehicle. The concepts of directional link and directional membership link were defined and an abstractive link segmenting model with clear direction and exact entrance was developed, which would be applicable for all kinds of roads in urban area. Furthermore, a method for data organization of the directional links also designed.

The algorithms of direct method and indirect method for estimating link travel time based on GPS equipped floating car are achieved using the programming language and verified with measured GPS data based on part road network in Changchun, and the characteristics and applicability of direct method and indirect method are compared using link travel time data acquired. The results show that the direct method and indirect method hold different characteristics and applicable scope respectively.

According to the analysis and forecast for the intention characteristics of resident trip,the survey for ms with the SP survey and orthogonal design were designed and the survey results were obtained.By studying the characteristics of resident trip in Changchun City,a MNL model of the resident trip mode choice was built up,and the impact of time and cost on the various types of traveler was given.Somere commendations about bus priority policy were proposed,the rate changes of traffic mode sharing was forecasted after implementing the policy,and finally,the reasonable structural model of residents trip in Changchun City was summed up.

This paper,based on environmental protection,determines the capacity of urban cars in aperiod by taking into accountresidents’different choices of trip modes on different distance and a minimum of travel time as the optimization goals.The method has been applied to Liaoyuan City and the calculated result proves that the model is validand has actual reference value,which will provide a basis for the carm an agement and Comprehensive Traffic Planning of the city.

In order to estimate the driver’s sight distance(SD),which changed with the landscape of crooked highway in mountain area,there lation of road alignment,landscape attributes and vehicle characteristics was analyzed,then the calculation of SD was simplified based on some assumption.According to there lations the SD estimation model affected by road landscape of crooked highway in mountain area was established.The SD model will act as a theory too lf or assessing the safety of crooked road in mountain.

Based on characteristics analysis of passenger flow and service state of inner facilities in passenger terminal, measurement of facilities capacity for each type of facilities was determined. Reasons were analyzed in a quantitative view which two phenomena brought. Furthermore Indexes for individual and whole adaptation analysis were offered under the precondition to satisfy smooth passenger streamlines in terminal. While simulation system of subway passenger terminal was also established with the object simulated as passenger streamlines, individual and whole adaptation analysis were hereby performed respectively for individual facility and all facilities. Results show that parameter selection and simulating optimization for facilities allocation can be accomplished by means of system simulation integrated with measurement of facilities capacity and indexes of adaptation analysis. Some reasonable improvement approaches are represented to enhance adaptability of facilities capacity and its decision basis is thereby provided for the operation management and optimal allocation of facilities in terminal.

This paper from the view of services design, driven by client needs, brings forward a new method that evaluating and improving logistics services ability. from the idea of a dynamic development, introduces  the QFD theory to the field of logistics service capabilities design. Develops a dynamic QFD model and HOQ linear programming model, taking example for automotive logistics enterprise, the paper carries out a demonstration analysis, and provide theory method for enterprises to work out logistics quickly responding to future market developments.

The theoretical model of composite beam was put forward for the concrete structure strengthened with Carbon Fiber Reinforce Polymer(CFRP) sheets based on state space theory. The state space model was established to calculate the deformation, strain and the interface shear stress of concrete structure strengthened with CFRP sheets, and the calculation results are in good agreement with the result of tests. The results show that the state space model of the concrete structure strengthened with CFRP sheets is rational after abandon the planesection assumption. And the delamination failure, rupture of CFRP and crush of concrete can be forecasted during the construction of strengthening concrete beams with CFRP loads. Meanwhile, the proposed model can provide a new design idea for the concrete structure strengthened with CFRP sheets.

The research on the vertical distribution of air voids of porous asphalt mixture was carried out using XCT technique, digital graphic processing technique and the fractal theory. The results show that the void features images are able to describe the connectivity and distribution characterization of porous asphalt mixture and there is no inplane connectivity in it; the air voids of the six sections studied ranges from 17.7% to 22.2%; the greatest discrepancy of air void quantity among different sections is 16, accounting for about onequarter of the total quantity; all the air void profile fractal dimension are larger than the corresponding air void area fractal dimension, and it has nothing to do with air void quantity. The air void gradation figure shows that the gradations in two orthogonal crosssection mostly distribute within 4~8 mm in the same sample, but with different distributional pattern. The contrastive analysis find that the discrepancy of maximum diameter between transverse air void and vertical air void is 2 mm with relatively close equivalent diameter.

This paper puts forward the diagnosis method based on fuzzy mathematics, introduces fuzzy mathematical method into the suspension fault diagnosis, and analysis the ambiguity relationship between the fault symptom and fault reason, establishes a fuzzy comprehensive evaluation model of the suspension fault diagnosis. And by means of the state parameters method, set up the membership function model and evaluate the suspension system comprehensively according to the Maximum Subordination Principle. The typical examples show the accuracy and practicality of the model.

The evaluation method of the overall detection ability of Automobile Detection Line(ADL) is put forward based on analytic hierarchy process and fuzzy comprehensive evaluation，then weights of each floor are determined, the index system of fuzzy comprehensive evaluation of ADL overall detection ability is established. Through researching the case, the process and method of evaluation is discussed in this paper. The evaluation system provides a quantitative evaluation method for ADL quality certification, the application is effective.

This article is based on a PXI bus, PXI bus, introduced in line with the standards dedicated multifunction board. PXI multifunction board dedicated primarily to complete a variety of signal input, output and control functions, including A/D converter, D/A converters and I/O function modules such as switch circuit, can effectively function on the closedloop control module, fully able to meet a wide range of systems and a variety of input and output signal processing requirements. Used in vehicle performance testing instrument systems and combined with CCD image acquisition and DSP processing system，it has the advantages of good realtime detection, high precision, high universality, fullfeatured, costeffective, easy to operate and so on.

In order to develop the automotive engine coolant hose products and detect its performance, by simulating actual condition, at which the coolant hose work, coolant hose highlow temperature vibration test system was set up. First of all, according to the actual vehicle driving conditions, the engine coolant hose vibration test system designed. and then place the test device into test cabin environment to proceed vibration test in different temperature at -40 ℃,23 ℃ and 70 ℃ respectively. The vibration test results of the hose outside diameter of the largest change rate at 23 ℃ is 15.96%, -40 ℃ low temperature is 17.68%, 70 ℃ is 19.22%.Experiments show that the high, low temperature vibration test is more suitable for product development, design and formulation of standard parameters.

According to the analysis of steering kinematics model and the parking environment built in SIMULINK, maneuver process of parallel parking was simulated. In terms of fuzzy logic control theory, steering controller was designed, which help the car to complete parallel parking automatically. Then the car controlled by the controller built in SIMULINK finished the backup course of parking satisfactorily. The simulation results demonstrated that the steering control arithmetic based on fuzzy logic control theory is not only simple but also responses quickly, providing a feasible way to parking assistance.

The modeling theory of UniTire and the application of UniTire in vehicle dynamic simulation software has been investigated in detail，by doing this, the application of UniTire in general dynamic simulation software ADAMS has been fulfilled. Through the simulation and experiment, the accuracy and the simulation ability in limited case of UniTire have been validated.

Presented the traction control strategy of a Full-time four-wheels-drive vehicle with a continuously variable transmission relying on controlling the transmission, inter-axle differential and brakes. Built a vehicle transmission system software inthe loop simulation platform with Matlab / Simulink software. Simulation to the proposed control strategy and the traditional control strategy relying on adjusting the engine, inter-axle differential and brakes in the low attachment coefficient road respectively. The results showed that the control strategy with the transmission participating significantly improved the vehicle road surface traction in the low attachment coefficient road, and was superior to the control strategy of relying on controlling engine.

The models of simplify realtime dynamic for the heavy vehicles was compared, and a 5DOF model of heavy semitrailer was selected out because it can better predict the future rollover risk. An algorithm of realtime TTR rollover warning based on the model and optimal active control strategies for antirollover LQR based on the algorithm were researched. The heavy vehicles were active controlled for rollover with differential braking. Accuracy and reliability of warning algorithms was also verified using TruckSim and MatLab Cosimulation for fishhook etc. steering situation. The results show this method can reduce rollover performance index of Lateral Load Transfer Ratio effectively and avoid occurred rollover and improve the roll stability of heavy vehicles significantly.

Based on the theory of vehicle system dynamics and applying the simulation software TruckSim, the characteristic parameters model and the corresponding aerodynamic coefficients curve model of a certain Chinese-made truck are established .The process of crosswind effecting on the truck running in a straightline is simulated, and the aerodynamic characteristic curve under the conditions of openloop is obtained. Using the software TruckSim ,the driver model is imported. Changing the parameters such as the wind speed, the wind heading and the speed of truck ,it is analyzed the influence of crosswind effecting on handling and stability of truck which is driven in a straight-line.

In order to adapt to new modern car styling, the aerodynamic numerical simulations based on computational fluid dynamics(CFD) are applied to the process of car styling. At first it was introduced that the process of car styling in this paper. Virtual wind tunnel test is one method of the best modern ways of aerodynamic design which will be applied in the fields of aerodynamic research widely. Together with one concept car design，the details in all the process of aerodynamic numerical simulations are narrated. Applying the virtual wind tunnel test aerodynamic drag coefficient, velocity contour and pressure distribution are got. Some advices to reduce aerodynamic drag of the designed car were put forward. In one word，it is one fast, simple and effective way to design aerodynamic car shape.

Taking a heavy truck cab as the example, a model of  the spotweld space based on finite element  was built to analyse the torsional stiffness of the model, and the amount and layout of the spotweld using the discrete variable topology optimization were optimized. Comparing with the original model in welding spot's number and the torsional stiffness, the optimized model has less number of welding spot, but the torsional stiffness of the two model is almost the same. So the method of topology optimization has guiding significance and great value on the spotweld layout.

In order to determine the modal density of complex sub-structure such as auto body, it was studied from theory calculation, test measurement, numerical analysis and virtual test aspects. The influence of the complex sub-structure modal density to vehicle interior noise obtained by method of theory calculation and virtual test was compared and analysed, according to the established cab Statistical Energy Analysis(SEA) model. The results show that the virtual test method used to determine the sub-structure modal density can improve the prediction accuracy of the SEA model.

Interior noise samples in the assistant driver ear's position from four types of cars at different gears and velocities were selected as the evaluation objects. Subjective evaluation test of sound quality was carried out with the paired comparison. Several primary psychoacoustical objective parameters of these samples were calculated. Applied the binaural sound pressure level spectrum synthesizing method to weight the left and right ear loudness, the correlation coefficient of the binaural loudness and sound quality subjective evaluation result are higher 0.107 than that of left ear loudness, proved that the binaural loudness objective evaluation model is more nearly the subjective evaluation result. By way of correlation analysis and multidimensional regression analysis, the equation between subjective evaluation level and psychoacoustical objective parameters was gained. The research results indicate that vehicle interior sound quality preference under the stable state condition is affected by two psychoacoustical parameters of loudness and sharpness.

Two-dimensional solidliquid melting phenomenon of frost on outer windshield surface with air impinging was simulated based on standard kε turbulence model by Fluent. The efficiency of windshield defrosting is affected by jet angle and sweep angle which determine the fluid flow patterns, average external temperature of windshield, liquid fraction and average liquidphase rate. The results indicate that the effect of jet angle and sweep angle on defrosting is nonmonotonous, and there is optimal scope of jet angle and sweep angle. The optimal jet angle is 60°and sweep angle is 10°under promissory conditions.

The thermodynamics models of direct air cooling system and CCACS was made, and the performance of CCACS was analyzed detailedly in this paper. Whereafter the income causation of the CCACS was obtained based on characters of the indirect air cooling system. It is very necessary to understand accurately the CCACS.

Composites based on thermal liquid crystalline polymer (TLCPa) and polyamide 6 (PA6) were prepared using a solutionmixing technique in this study. Differential scanning calorimetry (DSC), thermogravimetry analysis (TGA), wide angle Xray diffractometry (WAXD) and polarizing optical microscope (POM) were applied to study the compatibility, thermal stability and crystallization behaviour. The addition of TLCPa enhanced the thermal stability of PA6. The degradation onset temperature of composites with 30% TLCPa in weight increases nearly 27 ℃ rather than pure PA6. This is due to the increase of miscibility induced by the intermolecular interactions between these polymers. Crystallization behavior revealed that the TLCPa inhibited the orientation of polymer chain resulting in the decreases in crystallinity, but no effects on crystal modification manifested also by XRD spectra. XRD spectra of amorphous state appeared when the addition of TLCPa over 40% in weight.Crystal morphology refinement was observed by polanized optical microscopy(POM).

Holes are divided into singleface hole and multipleface hole whether boundaryloop of the hole is in one face:Singleface hole is be found by using topological relations and outerboundary & multipleface hole are distinguish by comparing volume of the bounding box. Singleface hole(s) of feature modeling was filled by deleting auxiliary feature. Singleface hole   which the feature modeling information is not existed or lost was filled by using bicubic Coons surface as multipleface hole. Hole(s) was filled accurately rapidly by algorithm, and provides necessary precondition for follow contact algorithm in numerical simulation of sheet forming process.

By numerical simulation of complex drawing and reverse drawing in one procedure forming of cylinder part, this paper analyzed how main process parameters affected complex drawing and reverse drawing in one procedure forming of the cylinder part. then based on the results ,a complex drawing and reverse drawing in one procedure mold for a kind of engine oil filter shell had been designed, and process experiments had been done. The process experiments achievement not only validated the reliability of the numerical simulation, but also applied the complex drawing and reverse drawing in one procedure forming technology into the production of the engine oil filter shell.

Based on the same stress and strain state between stamping edge tensile and simpleten siletest,amodified simple tensile test was proposed to researchon effect of sheared burronde formation capacity of high strength steel stamping edge.The test shows that the trend of the deformation capacity of stamping edgeis decreasing with anincrease of burr height.It is easily explained the reason why splitting on stamping edge appearing before deformation reachesits normal deformation capacity.The numerical analysis approach for spliton stamping edge was proposed to solve the problem of the traditional method failing to predicted gefractureaccurately.Furthermore,the specialnumerical analysis software was researched and developed.It is confirmed that the numerical analysis approach is both feasible and effective by analyzing on automobile exhaust pipe support flanging process.

For three-phase large welding inverter power supply, three-phase boost topology is introduced as the PFC scheme. With the merits of space vector modulation: digitized control，harmonic decrease，high ratio of voltage used，fast dynamic response and so on，and at the same time deadbeat and DSP control is introduced, as well as two-loop control of the system，we gained power factor is near unit and current THD is well below 10% in our digitized arc welding power system.

A new method of improved Butterworth filter was proposed to denoise feedback vicecurrent of inverter effectively and make currentmode PWM controller work fast and steadily in the environment with strong interference, high voltage and large current. First, the signal was analyzed with fast fourier transformation to get the frequency spetrum of useful signal and noise. Then IIR Butterworth filter was designed after analysis of processing speed with FPGA realization. In order to improve the start and tracking speed, moving average median filter was proposed. The simulation and experiment results show the filter is characterized by small phase shift and fast dynamic response. All digital inverter can work accurately and reliably.

Based on the bimorph piezoelectric vibrator, a piezoelectric pump with two active valves was designed and manufactured. Considering the mechanical structure of piezoelectric pump, the working principle of piezoelectric pumps with active vavles were analysed in detail. According to the requirements of piezoelectric pump on the drive signal, the test rig of piezoelectric pump with two active valves was built up. Finally, the performance of the designed pump was tested. The results show the maximum flow of the designed pump is 180 ml/min and the maximum reversal flow of the designed pump is 170 ml/min , the maximum output pressure of the pump is 6 kPa at the frequency 40 Hz.

The modal analysis of cantilevertype NC assembled frame for processing assembled camshaft was made by finite element method. The first ten steps natural frequencies and corresponding vibration modes of the frame were obtained, and each mode was analyzed and evaluated. According to the analysis, a modified structure design scheme for the bench and column of the frame was put forward to enhance the rigidity and lighten the weight, and the modal analysis of modified scheme was also made. The results show that the modified scheme can improve the vibration characteristics and lighten the weight of frame.

Based on the failure rate function of repairable system collectivity, a new failure rate model that focuses on system unit is presented by this paper, and an example is used to show the operation process. Both the theory and application show that this model could precisely describe the reliability state of system unit in any time of its life time. More over, it has great value in reliability predication and forming maintenance strategy.

Based on the statistical analysis of the failure data which were collected from some CNC lathes in the scene, it will establish the directed causality diagram of the internal subsystem and between subsystems. Considering the fault severity, frequency and the correlation coefficient between failure modes, it will obtain the failure related indicators of the subsystems. A critical subsystem that influences the reliability of NC lathe was presented, which provides the research direction for improving the reliability of NC lathe.

his paper proceeds with the existed limitations analyzing as while DBR was used to solve the uncertainties in remanufacturing system and takes the end of life engine remanufacturing system as an example. On the basis of qualitative analysis in the aspects as definition, calculation model and applicable conditions between inventorybuffer and timebuffer control mechanism, their implementation effects were compared and analyzed by simulation experiments. Simulating experiments show that two kinds of buffer control mechanism are of high goal consistency, i.e., they both enable Resource Utilization of bottleneck (RU) and System Productivity (SP) increased, average Work In Process (WIP) in front of bottleneck (BW) and WIP in total process (TW) decreased. From the perspective of DBR principles and its goals, the latter has better performance than the former in the following three aspects as buffer effect, objective realization for management of system and controllability of production planning.

Study the batch by coordination of one supplier and one manufacturer in supply and procure in the stage of mass production, and the coordination scheduling of one manufacturer to several customers in production and distribution in the stage of customization base on the customer order decoupling point, analysis the conflict between the supplier and the manufacturer, the manufacturer and the distributors caused by the traditional scheduling, furthermore, modeling the problem with nonlinear programming, finally, The conclusion that the cost of coordination scheduling less than the cost of the scheduling without coordination, and the scheduling by coordination is better than the traditional scheduling is given from the numerical example by simulation.

A new lattice Boltzmann model for simulating the seismic pressure wave is proposed. By using the ChapmanEnskog expansion and multiscales technique to the pressure distribution function, we obtained the moments of the distribution function of the seismic pressure wave equations. Using this model, we simulated the pressure wave motion caused by the multi-seismic source in the isotropic medium and single seismic source in two isotropic media with difference parameters. These numerical results reproduce the classical numerical results. By comparison between these results and results of the finite difference method, we find that the numerical results agree well with the finite difference method ones.

The speciality that the Escherichia coli can ferment lactose to produce acid which can produce depositions and make the indicator's color changed was utilized, an automatic detection system for enumeration of Escherichia coli in food based on color feature recognition technology was designed and the color change degree after incubating for 16 h was used to judge the Escherichia coli counts in the detection samples. The system can automatically extract the H、I、S color vector to act as the input vector and transfer the trained BP neural network to get the Escherichia coli counts. The experiments indicate that comparisons of counts of Escherichia coli by the rapid automatic detection system and by the traditional method closed correlated. Moreover, by using this rapid detection system, the Escherichia coli in food accurately enumerate within 18 h, which is much shorter than the 6 days detection time by the traditional method, which will greatly improve the products sale quality.

The model of multicenter extraction was brought forward on the base of Fick law. A mapping model of extraction time and concentration were set up according to the experiment. The extraction mechanism by the high intensity pulsed electric fields (HPEF) was discussed. The advantage of short extraction by the HPEF was explained in the abstract though the mathematics and geometry model of extraction. An effective theoretics analysis tool was supplied to extraction of solid and liquid matter though this model.

This paper researched on beer waste yeast aimed at getting trehalose which extracted by distilled water. Contrast analysis and multiple linear regressions were used to optimize the technology parameters by microwave assisted extracting in beer waste yeast. The optimal technological parameters were dosage liquor rate at 1∶30, microwave treatment time at 3 min, microwave power at 600 W, diffusion temperature at 100 ℃,diffusion time at 3.5 h, and the yield of trehalose reached up to 9.24%. Engineering regression model of extracting trehalose from waste beer yeast was set up.

In order to develop alkylpolyglycoside(APG)which is a highpurity and environment friendly biosurfactant,the processparameters of APG synthesis catalyzed by ACA microcapsuleim mobilized enzyme were optimized through single factor experimentation and twice or thogonalrotating combination igntest,andthe model of glucose conversion rate was studied.The results show that factors which have influence onglucose conversion rate are:reaction time,amount of adding water,reaction temperature and amount of alcohol,what’s more,the formeris better than the latter;reaction temperature is more interactive with a mount of alcohol,amount of adding water,reaction time,a mount of alcohol and amount of adding water is the same;the best synthesis process parameters are as follows:reaction temperature at 45 ℃,amount of dodecanol of 8 mL,immobilized enzyme of 5 g,amount of glucose of 1 g,without adding water,reaction time of 5days and the maximum of glucose conversionrate is97.9%.

Keeping aneyeon the projection sequencein traditional algebraicre construction algorithm,the proposed algorithmdoes projection in an order which is decided by the value of pseudoerror of each hyperplane computedin prior.So thegivenorderintr aditional algorithmturns to bean adaptive one.Experiment alresults indicate that the modified algorithm gives much better results than the classical one.

By exploiting the macroblock correlation in spatial and temporal scalable video coding, a fast mode decision algorithm was proposed. It chooses a part of the used search modes from the co-located reference macroblocks to compose the candidate group for the current macroblock. This will extremely reduce the calculation amount of motion estimation in enhancement layers. According to the experimental results, this algorithm decreases the encoding time with the negligible coding loss and bit-rate increases.

A frequency estimation method was presented for weak sinusoidal signal embedded in chaotic‘noise’based on parameter variation chaotic synchronization which one filtered driving variable isused to drive response chaotic system.Because the error data series between drive and response signal has only little chaotic interference and has the sinusoidal signal.Usingthecrossspectral MUSIC method to estimate the sinusoidal frequency in the error data.The method issuit for the situation where noises are mixture of strong chaotic noise and white noise.

The two levels search model is put forward,the minimum syllable step search algorithmis adopted for speech detection in this model,and then the binary search algorith misused to determine the exact position of speech segmentation.The theoretical analysis and practical operation results show that the searching efficiency canim prove70%than the traditional end pointdetection algorithm.This algorithmrequire good recording environment,it can be used for talking publications speechautomatic segmentation and speech end point detectionin good SNR.

For the shortcomings of huge calculation in MRI images segmentation with standard Fuzzy CMean algorithm (FCM), a new algorithm combined with histogram statistical information of Improved Fast Fuzzy CMean algorithm (HFKFCM) was proposed. Firstly, the method of multiscale window traverse is used by this algorithm to find the histogram peaks. Then these peaks are defined as the fuzzy clustering initialization centre. Meanwhile, the fast FCM method based on histogram statistic is used as ergodicity to reduce each iteration calculation. The simulation results show that compared with standard FCM algorithm and other improved algorithms, the proposed algorithm can be improved significantly in fuzzy and clustering effectiveness.

In order to make intelligent video surveillance system not only have a high accuracy rate but also be assured the real-time, A practical moving target detection and tracking algorithm was implemented.First,preprocessing the captured image in order to provide the best image data.Then implementing the forground/background segmentation using the method of Gaussian Mixture Background Modeling,extracting forground objects using morphological operations.Finally implementing target tracking using distance of space-based method.Through applying the algorithm to the experimental system,the paper came to a conclusion is that the system can detect the moving object and tracking correctly while ensured the real-time.

In E-learning, in order to satisfy users' personalized demands for learning resources, in this paper we present a learning resources personalized recommendation algorithm based on semantic web technology. First of all, based on users' evaluations and browsing behaviors of the learning resources, we define the learning resource set and core concept set for a user; secondly, based on the relations of concepts in domain ontology, we obtain the user's preference similarity by computing semantic similarity between the different users' learning resource sets and core concept sets respectively. At last, we recommend collaboratively the learning resources according to the user's preference similarity. In addition, the learning objects for management of learning resources can reduce the complexity of similarity computation. We applied this algorithm to the personalized learning resources recommendation system, the experiments show that the algorithm improved the effectiveness of the recommended learning resources, especially for new resources and new registered users.

The advantage of hybrid programming of C++ and MATLAB is both the powerful function of character operating in C++ language and the convenience of matrix treatment in MATLAB. As the result, the mixlanguage programming is applied in structure reanalysis using combined approximations method. It is efficient and convenient to reanalyze the structure by running the application in ANSYS, thus the secondary development of ANSYS is achieved. At the same time, it provides a powerful approach to reanalyze large-scale structures.

A feedback model was put forward in order to improve the realtime capability of intelligent video surveillance. First, the background subtraction method based on Gaussian mixture model was adopted to detect moving objects. And the Kalman filter algorithm was adopted to perform the realtime object tracking. Then a feedback model was put forward. In the model, the results of tracking was fed back to the detection stage in order to restrict the detection regions. By experiments, the proposed method(which used the feedback model) was compared with the common method(which didn′t use the feedback model). The experimental results illustrate the good performance and enhanced realtime capability of the proposed method.

A new method for real-time controllable detail synthesis and rendering of large terrain based on data chunk was proposed in order to overcome the difficulty in accessing the dataset for rendering the real large terrain, and reduce data storage and management pressure. Chunk hiberarchy is adopted to organize samples, and the controllable detail synthesis presented is chunk-based. Chunk subdivision is used to generate new samples and a Perlinnoise based controllable multifractal algorithm is given to compute details. The influencing factor for the fractal makes sure the details generated are controlled by the real terrain features. A uniform refine criterion is studied to integrate detail synthesis and the view-dependent LOD rendering mechanism. The new chunks synthesized together with original chunks form a two-level adaptive hierarchy, and large terrain rendering is completed. The method is implemented in a GPU-oriented way, and real-time synthesis algorithm is achieved on GPU. Experiments show that the method can real-time produce fine large terrain scene similar to the true terrain using limited terrain samples.

The wavelet neural network(WNN) is applied in many fields for its advantages of fast convergence，simple structure and computation since it has been presented. According to its strongly approaching ability to single dimensional signals, the Grey Relevancy Analysis theory is adapted to analyze the characteristic parameters and damage index of structural damage. The obtained relevancy degree of parameters is regarded as characteristic parameters of dynamic structure, and is transformed into weight value of singledimension input. Furthermore, the WNN model is presented with the Grey Relevancy Analysis. In order to improve the convergence ability and approaching effect, the study algorithm of WNN is ameliorated by local study policy. The simulation result indicates that the proposed method is superior to traditional Neural Network model.

The effects of over voltage, over current and surage impacts on the dynamic performance of induction machine were studied.The performance indexrepresentingthyristor’s power loss was established by means of state space modeling tool.Taking advantage of impacts’restriction conditions,the idea that obtaining optimal parameter configuration sthrough optimizing performance indexinsteady state was proposed.The simulation curves and actual wave forms show that this method significantly restrain simpacts and reduces power loss.