Search

As a key component of bus dwell time, passenger movement time in boarding and alighting processes plays an important role in affecting service reliability and service quality. Boarding and alighting processes are impacted by several factors, including vehicle configuration, friction between boarding and alighting passengers, bus floor level, and fare payment methods. The crowding effect refers to the interaction between passengers boarding, alighting, and on-board. In this paper, we introduce two indices, namely rank and occupancy, to describe the crowding effect. Rank reflects the sequence of passengers boarding and alighting, which reveals how passengers’ movement time changes in boarding/alighting process. Occupancy indicates level of crowdedness in vehicle, and quantifies the friction from on board passengers during boarding and alighting process. Based on time-stamped smart card transactions, we measured the inter-tapping interval between two successive boarding (alighting) passengers to identify the effect of rank and occupancy in determining boarding/alighting processes. Before our analysis, a cluster based outlier detection method is applied to guarantee data quality. Using large quantities of boarding/alighting observations from smart card data, we found that both rank and occupancy have strong effects on boarding and alighting processes. For occupancy, 60% is a critical threshold for boarding process, after which a dramatic increase of average boarding time is observed. This research may help better understand the passenger behavior in boarding and alighting processes.

The application of network science has greatly improved understanding about the structure and properties of public transport network (PTN). Using tools and methods provided from network science, several standardized representations have been developed and analyzed to understand the topological structure of PTN from different aspects. In this paper, the authors present an empirical study on the bus network in Singapore. Different from previous studies based on a direct mapping of PTN on road networks, they focus on the weighted C-space representation of public transport networks to better characterize the interaction of different bus services. By exploring the statistical properties of this representation, they find that the bus network in Singapore exhibits similar patterns as many other cities in terms of distributions of edge weight, vertex degree and vertex strength. However, a superlinear scaling is found when the authors examine the dependency of a vertex’s strength on degree, indicating that PTN in Singapore provides people with more alternatives to travel and transfer. The C-space representation also enables us to further explore the interactions between services by examining the community structure and the corresponding reverse mapping on road networks. This helps us to identify those crucial ‘bridges’ in a city, which is shared by different communities and connect diverse regions.

In order to realize the intelligence of a transportation infrastructure management system (TIMS), deterministic models are built according regressive methods, and the pavement performance model is a classic, using the Bayesian approach, which introduced a prior distribution to describe the related information of initial parameters. The new collected data reviewed a sample of the population, then according to the prior distribution and the sample, deduced a posterior distribution that was used to update model parameters. In the next stage of the model updating process, the deduced posterior distribution was a prior distribution and repeated the above process, and model parameters were updated consistently. This Bayesian approach was applied to the case of parameter updating of pavement performance prediction, and it was shown to be a mathematically valid methodology to deal with the problem of the parameter updating. This approach is easy to program, which can be used to realize intelligence of TIMS.

Aggregated total demand is often involved in microeconomic public transport models to determine the optimal frequency and bus size on a single bus route. With the implementation of Automatic Passenger Counter (APC) systems, massive boarding and alighting counts are available for analysis. This study proposes a microeconomic model to determine frequency and bus size using APC data. The proposed model considers the influence of crowding by specifying the dependence of the value of travel time on bus occupancy. It is revealed analytically that the model performs equally well as a model that uses origin-destination (OD) flows as inputs and that ignoring the influence of crowding would result in underestimating frequency and over estimating bus size. Numerical results suggest that a model based on bus load and a given bus size underestimates frequency, and a model that uses the total demand as input tends to underestimate frequency and overestimate bus size.

Park-and-ride (P&R) is a conventional way of transportation demand management that can relieve urban traffic congestion and parking pressure. Nowadays, the total usable capacity of P&R parking lots in Shanghai, China, cannot satisfy the demand. In this research, a user travel behavior survey was designed based on this phenomenon, the factors affecting travel behavior were analyzed, and a binary logit model was constructed considering both traffic congestion and subway congestion factors as well as models of roadway congestion and a user transfer model implemented separately. Travel time was the most important output factor in the model, especially for highly congested road segments. Fees also affected the choice of travel mode when roads were less congested. Those who had lower incomes, lived in suburban areas, or had non-Shanghai vehicle licenses preferred to use P&R.

Heavy vehicles have significant impact on the traffic performance of freeways. The passenger car unit (PCU) is a common measure used for freeway design and analysis. With the availability of a huge amount of empirical traffic data, investigations on the determination of the PCUs are feasible. This paper uses the simple linear density-speed function to estimate the capacities and investigate the relationship between capacities and the percentage of heavy vehicles. Empirical traffic data from five locations from the Outer Ring in Shanghai have been utilized for the analyses. The results show that the PCU value decreases with the rising percentage of heavy vehicles when the percentage of heavy vehicles is larger than 20%. In cases where the percentage of heavy vehicles is smaller than 20%, the PCU could be either increasing or decreasing, and is strongly dependent on the capacity that is estimated when there are no heavy vehicles in the traffic.

Existing ramp metering strategies mainly focus either on local control applications or on the coordination of ramp metering along the corridor. There appears a lack of effective indicators for identifying the control timing and evaluating the effects of ramp metering. Macroscopic Fundamental Diagram (MFD), describing the relationship between productions and vehicle accumulations in a traffic network, is useful in representing the traffic states of a traffic network. Traffic breakdown is a stochastic phenomenon and can be used for analyzing the instability of traffic flows. This paper proposes a novel performance indicator, namely the traffic performance index (TPI), for determining the control timing of ramp metering on urban freeway corridors based on the MFD and traffic breakdowns. It is found that the determination of the control timing of ramp metering on urban freeway corridors should consider the tradeoff between the network accumulations and the stability of traffic flows. It also shows that there exists a critical value of TPI, at which the network production is high and the traffic flows are relatively stable.

The road authority targets good ride quality in its pavement management. The International Roughness Index (IRI) is the most common way to describe road roughness, which is defined by means of quarter car model. For urban road with short section, complex traffic conditions, IRI calculated by a one-way vertical section is inappropriate to evaluate the pavement roughness. In this paper, IRI data detection by auto device was analyzed to make sure it is not influenced by the detection measure. The automatic device used to measure IRI in this paper has five laser probes. Detection position has significant impact on IRI data. IRI is processed to take in and consider the whole transection. Also proper road section length should be studied.

Line-structured light pavement detection has been developed for about two decades, but it is not yet mature because of unstable measurement results. An image-processing procedure based on ROI (region of interest) images is proposed in this study, which includes slight center extraction, connection, smoothing, and the rutting parameter extraction. The light center extraction is the foundation of the entire process. Two extraction methods are discussed in this study, Gaussian method and the gravity method. The Gaussian method can find the light center curve precisely, but the calculation process is time consuming. The gravity method is based on the assumption that the light intensity of the laser line distributes uniformly along the cross-sections, which is a simplification of the Gaussian method and has a high processing efficiency. The light center curve extracted by Gaussian method is considered as the true value to be used to evaluate the accuracy of the gravity method. The result shows that the error introduced by simplification is acceptable, and the gravity method can be used in practice instead of the Gaussian method.

The shear stress of asphalt layers caused by wheel load plays a crucial role in rutting. A finite model of asphalt pavement was built and verified the validity. By comparative analysis, the median line perpendicular to the loading rectangle long side on loading surface was the most shear stressed section. A proposal was put forward that the average of shear stress for every sub-layer under positions 3, 4, and 5 on the outer half of the median line was applied to be shear stress of this sub-layer. The shear stress curves from this method had a stable shape and complied with the objective law that the shear stress of upper and middle asphalt layer was relatively high. A formula was put forward to simplify the analysis process of shear stress and verified based on large-scale calculations and improved probability density function of Gamma Distribution.

Asphalt overlay on old cement concrete pavement means overlaying an asphalt concrete layer on the old cement concrete pavement to transform the old road. In this paper, a variety of factors that affect pavement comprehensive evaluation were analyzed and summarized to build a comprehensive evaluation system specific for asphalt overlay on old cement concrete pavement. The evaluation system contains a level one indicator, four level two indicators, and several level three indicators.

Coordinated signal control on arterial streets has been widely used and effective evaluation can help to optimize the progression scheme. As assessing performance and identifying existing problems are essential purposes of evaluation, an analysis of common traffic problems on the arterial is made and a qualitative relationship between traffic problems and signal control is established. Based on it, a problem- identification evaluating model is formed at two levels of intersection and arterial with multi-goals, under which related evaluation indices are chosen. Weights of the indices are determined separately under saturated and unsaturated conditions using Analytic Hierarchy Process. The method of fuzzy comprehensive evaluation is used and the membership functions of chosen indices are determined to get the evaluation model. This evaluation model is used in assessing an arterial in Shanghai

Vehicle trajectories provide valuable information of traffic dynamics. The paper proposed a modified video-based trajectory data processing approach to reliably and efficiently collect trajectory data on congested urban roads or at intersections. The approach is on the basis of Knoop’s idea of video transformation so that traces of vehicles are visible in a single image. Different from the existing method, a mode-based background subtraction algorithm is applied in the RGB color space to alleviate the influence of a small gap. Vehicle trajectories are then directly detected by searching connected components in the segmented binary vehicle trace image. Test results show that the complete process procedure of a single lane can be finished in 1 hour for a 5-min video. Almost 97% of the trajectories can be accurately detected which is very promising.

With the increasing interest in Intelligent Transportation Systems (ITS), the construction of traffic data collection systems in many cities has expanded in recent years. All these data have built a good basis for the urban traffic managers to better understand the real traffic conditions and future developing trends. Trying to use these data well in this way, Shanghai Traffic Information Management and Evaluation System (SHTIMES) has been developed in Shanghai, China. The objective of SHTIMES is to integrate all the collected data from different sources and give a reasonable traffic operation evaluation so that the corresponding traffic managers can make their decision based on that. Three key aspects during SHTIMES development are addressed in this paper: Firstly, in the data collection aspect, the “closing ring” scheme is adopted for the surface road network, which can infer all the traffic flow over the network; secondly, in the model part, a united evaluation model for all the urban roads is introduced; finally, a C/S structure is used in the system design. All the functions related to calculation and are assigned to the server end and all the functions related to the information management and analyses are assigned to the client end.

Recently, with the increase in traffic accidents, more and more attention has been paid to road safety. Traffic ergonomics is the focus on improving the coordination among people, vehicles and road environments from the angle of traffic participants based on ergonomics. In this paper, by the theory and method of traffic ergonomics, a variety of factors that affect highway safety were analyzed and summarized to establish a highway safety evaluation system based on traffic ergonomics. The evaluation system contains a level one indicator and four level two indicators, and there is a summary of the contents for each of the level two indicators. Finally, an accident-prone segment of Badaling Expressway was been taken for example. There were analyses of reasons for accidents and measures for improvement based on the highway safety evaluation system.

With the rapid development of road construction, the contradiction between urgent conservation needs and limited conservation funds is becoming more and more prominent. Traditional conservation management model is gradually replaced by the development and application of pavement management system. Relying on the requirement of road management department of Suzhou, the architecture of Network-level Decision Support System for Urban Road (NDSSUR) was constructed firstly. Then, a detailed analysis of the most important three parts of the system ― performance evaluation, maintenance decision and performance prediction were carried on. Finally, based on the local dynamic detection data, NDSSUR was applied to evaluate pavement performance, suggest large and medium repair and do a long-term planning for five years within the urban areas of Suzhou. Results from the program fit the actual condition of the road well.

Control delay is a key performance measure for evaluating the level of services of signalized intersections. The existing empirical data driven methods are difficult to determine the average control delay precisely and efficiently. This paper presents a semi-automatic processing approach to quickly measure average control delay using surveillance video data. On the basis of the vehicle trajectories obtained using video processing techniques, three critical points along the vehicle trajectories are extracted manually: the first two critical points are before the vehicle decelerates to determine the stable driving speed of the vehicle, and the third one is after the vehicle completed the acceleration to resume the stable speed of the vehicle. Using the coordination of these three critical points, the control delay for each individual vehicle can be estimated. The approach is verified in an intersection of Shanghai. The results show that the average control delay can be properly measured with relatively short computation time and low cost.

Automatic pavement detection based on video or image acquisition has become a common means of modern pavement management. However, the effect of pavement image processing is not quite well understood due to large amounts of noise. The low portion of crack pixels to the entire image makes the situation even worse. In this research, an efficient way of image preprocessing was developed for pavement crack images. This study included three steps, background correction, Gaussian smoothing and histogram transformation. Actual pavement crack images were preprocessed by the new method and then segmented by Otsu method. The results show that the preprocessing steps presented in this study dramatically dampened the impact of noise on image segmentation and retained most of the distress details at the same time. The Otsu method provided good segmented crack images if the images have been properly preprocessed. In the case of the crack images tested in this study, the preprocessing effect in the line cracking images was good enough for the following segmentation, and the effect in the alligator cracking images was even better.

The turning movement proportions (TMP) of urban road intersections play an important role in traffic planning, signalized intersection traffic control, and O-D estimation. This study analyzes a great deal of turning movement real data and adopts a MATLAB linear programming (NP) based approach with assort and range constraint equations to estimate the turning movement of intersections. Application of the NP based approach through a case study in Shanghai shows estimation results are in good agreement with field collected data. As a comparison, the NP model shows higher consistency with real data than Maximum Entropy Model or Furness Model.

In order to investigate the role that sulphur additives play in asphalt mixtures, the four-component analysis was used to examine the change of component of sulphur extended asphalt. Freezing and thawing test, splitting test, rutting test, and low temperature bending test were used to find how much the sulphur content effected the performance of asphalt mixtures. Test results showed that sulphur could change the group component of asphalt in some conditions, the 25°C penetration and 10°C ductility of asphalt was increased at first and then decreased with the increase of sulphur, and the softening point of asphalt showed monotonically rising with the increase of sulphur. Asphalt could only react with sulphur of a limited content of 15%~25% of asphalt, and the sulphur over consumption restrictions could not react with asphalt.

Performance of asphalt mixture incorporated with waste glass as a part of aggregates is studied. Results show that Marshall Stability and flow, dynamic stability, bending strength and strain of glasphalt mixture decreases a little; however the bending modulus, moisture susceptibility and light reflection improve compared with normal mixtures. Mechanical properties of glasphalt could achieve the engineering demands and glasphalt paving will bring great social and environmental benefits.

This paper discusses the embodiment of bounded rationality in travel choice. In classical utility functions based on hypothesis of fully rational consumers, travel time is assumed to be negative utility which is minimized by travelers, whereas the average travel time is observed to be relatively constant over the past 40 years. Thus, in order to further understand the incentives of travel choice, a survey was conducted in both London and Shanghai to collect data about individuals' perceptions of ideal travel time in relation to different travel purposes. The result implies that the bounded rationality characterize the probability distribution shape of ideal travel time which can be well fitted by generalized extreme value distribution.

Cracking is one of the main disease types of asphalt pavement. Currently, adding polyester fibers is one of the measures most commonly used to reduce cracking. Firstly, performance-enhancing mechanism on anti-cracking ability of asphalt mixture adding with polyester fibers was studied. Then, a typical polyester fiber was selected, anti-cracking ability of asphalt binder and asphalt mixture for different fiber lengths and different fiber dosages were tested. BBR test was selected to evaluate anti-cracking ability of asphalt binder while four-point bending beam test was selected to evaluate anti-cracking ability of asphalt mixture. Study showed that anti-cracking ability of asphalt was excellent when the length of polyester fiber is 6 mm, and the optimum dosage is between 0.2 ∼ 0.5% of the mixture.

The accelerating pace of urban development has given birth to the large-scale transportation pivot, in which abundant traffic is abstracted and congregated. Drivers often have difficulty obtaining reader-friendly available car-parking place information in this area, inversely been confused with complicated layout of car-park and having trouble finding their correct car-parking place when they returned. This paper describes a choice behavioral model based on principal components analysis to survey data. The model was used to determine the demand intensity of different car-parking guidance information, which influences design of the car-parking guidance system. A triple-level sign system was developed for indoor car-parking guidance aims. The procedures were applied to the car-parking guidance system of Hongqiao comprehensive transport hub in Shanghai.

Recently, the Next Generation Simulation (NGSIM) has attracted many researchers’ interests because of its representative data and rich contents. Based on the NGSIM data, this paper calibrates the multi-anticipative car-following models, including multi-anticipative IDM, multi-anticipative FVDM, MHVD model, and MVD model. Before calibration, this article pre-processes the accelerations and velocities since there are several defects in the original data. The calibration of car-following models and multi-anticipative car-following models is grounded on the Levenberg-Marquardt algorithm; furthermore, the calibration results of car-following models can provide important reference value for researchers, such as the appropriate estimating ranges of model parameters, while the results of calibrating multi-anticipative car-following models not only introduce the parameter estimation ranges but also testify that, in reality, the multi-anticipative models are more suitable and two preceding vehicles can be a better choice.

Considering circumstance characteristic and data availability of local area road network, the authors attempt to make a new model for traffic flow assignment based on simulation of driver's stochastic choice behavior in the intersection. The formulation for local area road network assignment problem is proposed on the assumption of random turning behavior. And the compute process of traffic assignment coefficient matrix for local area road network is provided in this paper. The new traffic assignment model has been used in traffic analysis for an actual urban road network. The results comparing surveying traffic flow data and determining flow data by the previous model verify applicability and validity of the proposed methodology.

In order to make the best of limited traffic detectors, know about the traffic performance of urban road network, and then set effective management measurements, the paper puts forward a loop detector-layout method called "close-loop", and a traffic assignment method for local urban road network. A traffic assignment model is founded based on the stability test of turning split of intersections by real data, and the uniqueness of assigned link flow is discussed. Monte Carlo Stochastic Modeling Method is adopted to design vehicle turning behavior module, and a simulation model is founded. The model shows better performance and applicability in a real local network, and can provide reference for founding related traffic models for Advanced Traffic Management Systems (ATMS) and the layout of traffic detectors.

Asphalt mixture is a typical temperature-sensitive material. Most former researches on evaluation of rutting performance of asphalt pavement were mainly focused on deformation in high temperature especially in 60 ° C. In fact, accumulated shear deformation of asphalt pavement will lead to rutting damage under changing temperature. In this paper, shear fatigue test under changing temperature was adopted to simulate this phenomenon. The result shows that shear deformation of asphalt mixture not only formed in high temperature, but also leapingly developed under changing temperature, which is similar to the real situation in site. Compared with typical models which simulate shear fatigue deformation of asphalt mixture, on the base of time hardening procedure, a new model of shear fatigue deformation of asphalt mixture under changing temperature is provided in this paper. Because this new model takes nonlinearity of fatigue deformation curve and sequence of temperatures into account, it can perfectly simulate test result. Shear fatigue equivalent temperature of typical areas were calculated by this model. The results shows shear fatigue equivalent temperature can display the reaction of air temperature to pavement rutting performance. Consequently, shear fatigue equivalent temperature can be taken to estimate pavement rutting performance in different climate.

Permanent deformation(PD) of asphalt mixture caused by high temperature (above 50°C), instead of the accumulative deformation caused by moderate temperature (20°C to 50°C), attracts more attention. First, distributions of pavement temperature based on continued hourly average pavement temperature data in Shanghai and Wilson County were introduced. The result indicated that percentage frequency of pavement temperature below 50°C is more than 90%, and that above 50°C is less than 5. Second, PD of two asphalt mixtures measured by APA (asphalt pavement analyzer), dynamic stability (DS) of four mixtures tested by wheel tracking device and shear strength of five mixtures tested by uniaxial penetration test were introduced. The results show that PD at 40°C is almost 50 percent of that at 60°C, and that the ranking of high temperature performance for the same mixture changes with temperatures. At last, result of rutting depths calculated by the Mechanistic-Empirical Pavement Design Guide (MEPDG) suggests that rutting depth for high temperature months is almost two to three times of that for moderate temperature months. Therefore, PD of asphalt mixtures should be evaluated considering pavement temperature distribution. INTRODUCTION Rutting of asphalt pavements is a serious concern for a long time due to the increased traffic and wheel loads in China. Thus, it is necessary to predict the PD of a mixture before construction. Several test methods are in practice to assess the PD. The commonly used test methods include diametral tests, uniaxial tests, triaxial tests, shear tests, empirical tests, simple shear tests and simulative tests which are basically accelerated laboratory PD prediction tests (Sousa et al., 1991). However, almost all the tests evaluate the PD at constant high temperature such as 60°C or PG (performance grade) high temperature, while PD at moderate temperature was ICCTP 2009: Critical Issues in Transportation Systems Planning, Development, and Management ©2009 ASCE 2793

In response to on-ramp queue length and control-threshold restriction, an ALINEA-based local traffic-responsive metering strategy controls the Wuyi on-ramp in the Shanghai urban expressway network. The ALINEA parameter study results, integrating data analysis, determined the ALINEA parameter’s values. The analysis of a local ramp’s traffic-responsive metering mechanism fixes on its control-threshold parameters. The Wuyi on-ramp data analysis determines the control-threshold parameters and builds the Wuyi on-ramp control-threshold table. The queue-length subsection restriction model was built according to the analysis of an ALINEAR classic queue model given the Wuyi on-ramp’s physical characteristics. The control-threshold and ALINEA algorithm and the queue length subsection restriction model constitute the Wuyi on-ramp local traffic-responsive metering strategy. The off-line simulation results show that this ramp metering strategy can not only even out the ramp volume’s pulse characteristics but can also increase the mainline speed and relieve the mainline downstream congestion.

Partial hourly pavement temperature variation measured in Shanghai was analyzed. Cumulative frequency of high temperature (above 50°C defined in this paper) was more than 90%. However, cumulative frequency of moderate temperature (20°C to 50°C defined in this paper) was less than 10%. Then permanent deformation (PD) of two asphalt mixtures at the temperatures ranging from 20°C to 60°C was measured using the APA (asphalt pavement analyzer). The results show that PD at 40°C can be as high as 50% of that at 60°C. Therefore, the PD property should be researched systemically. Firstly, the cumulative PD curve of a typical asphalt mixture was measured and defined by two zones. In the primary zone, the PD accumulates rapidly. The incremental PD tends to decrease, reaching a constant value in the secondary zone. Then, the PD property of asphalt mixture under varying temperatures was researched in laboratory. The result shows that the trend of the PD curves at the second stage was the same as but less than that at the first stage. Based on this study, it was concluded that temperature, number of load cycles, stress history, and intermission time are some of the major influencing factors on the PD of asphalt mixtures.

Relieving traffic congestion on roadways has always been one of the priority concerns of traffic engineers. Thoroughly understanding the factors which cause traffic jams plays the fundamental role in planning, design, construction, and management of roadway facilities. This paper focuses on experimental observations and theoretical analysis of urban freeway congestion due to complex features of traffic flow, shock wave and state transition. Real time dejected data were collected to calculate shock wave speed and establish vehicle trajectory models. A typical urban freeway segment in the center of Shanghai City is used in this analysis. As a result, conclusions have been drawn and the theory of shock wave speed could be used to help establishing traffic control strategies and to control queue length.

In order to meet the need of traffic route guidance and traffic monitoring system in urban areas, this paper introduces a method to estimate and predict traffic condition and travel time between arbitrary locations by utilizing the Extended Kalman Filtering (EKF) framework. The basic concept is to fist predict future traffic conditions using macroscopic dynamic models that are integrated with EFK based on fixed traffic detectors in urban areas, and then predict actual travel time using "fictitious car" method. Finally, a case study in Shanghai is presented. The results demonstrate acceptable applicability and precision method. This study may lead to better traffic control and traffic route guidance, and improve the accuracy of broadcasted traffic conditions.

The temperature distributions in asphalt pavement have a great effect upon the bearing capacity and performance of asphalt pavement. The Transportation Engineering School of Tongji University started comparative research about 5 years ago. Thermal sensors were installed in several pavement sections in different regions to monitor the temperature fluctuation inside the pavement structure. Based on data from this research program and related weather information, the application of SHRP and LTPP pavement temperature prediction model in China is verified. Then statistical temperature prediction models established through regression analysis are introduced in detail. Compared with the other existing prediction models, the models show comprehensive applicability and excellent accuracy.

Traditionally, the capacity of an urban street is defined for a single direction of travel as the capacity of the through movement at its lowest point (often at a signalized intersection). The results of analysis in this paper show that this is not reasonable. On one hand, the downstream intersection approach may not be the bottleneck of the whole street due to the possible existence of many affecting factors including signal density and it’s hard to judge the real bottleneck in practice. On the other hand, the capacity of a bottleneck cannot represent the street capacity effectively, and even might be a higher estimation under some circumstances. So the concept of Equivalent Capacity is proposed here, which can treat segment and its downstream intersection as an integrated system by introducing the travel speed as a variable. The primary evidence of the advantage of this new concept is given by a simple survey. Finally, the generalized definitions of flow, density and speed are used to explain the theoretic implication of equivalent capacity.

This paper discusses the structural design criterions of perpetual pavement in China. Three representative methods were selected to analyze the relationship between the threshold strain and the service life. Then, the design criterions were statistically analyzed according to the calculation of the representative perpetual pavements of USA. Given the dynamic modulus and the axle load of 80KN, the average tensile strain at the bottom of the HMA is about 64 micro strains, and the average compressive strain at the top of the subgrade is about 170 micro strains. With static modulus parameters and 100KN axle load utilized, the Guangzhou to Shenzhen Freeway in China with cement treated course was analyzed. The calculations reveal that the strain is smaller than the threshold strain of USA perpetual pavements. Further more, the perpetual pavement structural design criterions for China were put forward as follows: the tensile strain at the bottom of the HMA should be less than 120 micro strains and the compressive strain at the top of the subgrade should be less than 280 micro strains. A low stiffness is recommended for cement treated layer in a pavement. In the end, the perpetual pavement trial sections constructed in 2003 were used for checking calculations, which can satisfy the criterions of perpetual pavement of China.