Journal of the Eastern Asia Society for Transportation Studies
Online ISSN : 1881-1124
ISSN-L : 1341-8521
Volume 6
Displaying 101-150 of 314 articles from this issue
PARKING
  • Hsien-Ming CHIU
    2005 Volume 6 Pages 1344-1353
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    In this study, an application of the location theory on the allocation of the off-street parking facilities is explored. A multi-objective linear Integer Programming model is developed for this problem. Due to its linearity, the proposed compromise programming method for the model can be solved by the LINDO compute software. Based on the results of the numerical experiment, the proposed model can provide reasonable solution to the problem addressed. The following information can be provided by the proposed solution procedure for the decision makers to determine the locations of the off-street parking facilities: the optimal locations of facilities, the demands served by each facility, and the optimal type for each of the selected parking facilities.
    Download PDF (599K)
TRANSPORTATION FLOW MODELING
  • Sung Bong CHUNG, Ki Han SONG, Sang Yeon HONG, Seung Young KHO
    2005 Volume 6 Pages 1354-1367
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    In this study, the deceleration model and the sensitivity term were developed by considering actual driving behaviors to prevent rear end collisions in the car-following situations, especially in deceleration situations. Test vehicles capable of measuring the speed, acceleration, and spacing between the vehicles were used to collect data in the field survey. The deceleration model developed by linear regression has a high coefficient of determination (0.932) and for the goodness-of-fitness test of the sensitivity term, the chi-square test was performed under 95% confidence level. The results (χ2=59.74<χ20.05=68.24) showed that the sensitivity term has practical significance. According to the results, the factor that determines the drivers' characteristics in deceleration situations is the lapsed time until the driver starts to respond to the given stimuli. The results indicate that the lapsed time until driver reaction corresponds to the α value, which represents the drivers' characteristics in the GM model.
    Download PDF (719K)
  • Hyunho Chang, SeungKirl Baek, Seong J. Namkoong, Byoungjo Yoon
    2005 Volume 6 Pages 1368-1381
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    It's reported that real traffic flow has three phases such as free, synchronized, and stop-and-go flow, and that macroscopic flow-density relationship, called the fundamental diagram, has various features such as inverted V-shaped, inverted U-shaped, bell-shaped, reversed λ-shaped, and reversed λ- shaped with upward concave. Cellular Automata (CA) theory for microscopic vehicle simulation modeling is capable of simulating large traffic road networks faster than real time using both memory-saved and high-computing data structure. Although simplified CA models are able to robustly reproduce the basic properties of uninterrupted traffic flow, they have shortcomings to the description of some flow-density relation diagrams such as revered λ- shaped with upward concave and bell-shaped relation. Based on existent CA Car-following models, this paper proposes an improved CA model integrating two additional rules for both stopping maneuver in the tail of traffic jam and low acceleration within traffic jam. Consequently, the improved CA model describes microscopic vehicle behaviors such as braking maneuver at the tail of traffic jam and low acceleration within congested traffic flow more realistic than present CA models and is able to reproduce various flow-density relationships including revered λ- shaped with upward concave and bell-shaped fundamental diagrams that existing CA models are hard to reproduce.
    Download PDF (597K)
ROAD NETWORK MODELING
  • Richard CONNORS, Agachai SUMALEE, David WATLING
    2005 Volume 6 Pages 1382-1397
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    The Network Design Problem (NDP) refers to the optimization problem faced by a planner whose aim is to improve a transport network, drawing on limited resources. Though the NDP may lead to, for example, a set of tolls that maximise social welfare, often no consideration is made of the distribution of resulting benefits and costs across the population of travellers. We consider a network under probit stochastic user equilibrium (SUE) with elastic demand, disaggregated into multiple user classes with different values of time and link-specific tolls. We propose the Theil measure for quantifying equity, which can be incorporated either into the objective function or the constraints within the NDP. A sensitivity analysis of the SUE flows provides the basis for computing the Jacobian of the social welfare function and of the Theil measure. This allows gradient-based optimisation algorithms to be used in solving the NDP. Numerical examples are reported.
    Download PDF (650K)
  • Bahram MOJARRABI, Nikolaos VOGIATZIS
    2005 Volume 6 Pages 1398-1409
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    Statistical mechanics mapping techniques have been applied to improve the traffic delay times encountered around a large shopping centre in Adelaide, Australia. The shopping centre is large enough to act as a super node attracting a significant number of vehicles within the network.
    In the context of the locality scope, the dynamic evolution of the traffic network can be mapped as a Bose gas. We define nodes to be the sites at which a vehicle stops; thus can be represented as energy levels. Links are paths or inter-connectors that allow one or more vehicle to travel between them. They represent particles in Bose-Einstein picture. The mapping shows the emerging of the topological distinct phase-states: “fit get rich” and “winner takes all” when a fitness equation was introduced. The fitness parameter can then be used to control the evolution of these phase-states through its control parameter T. A Locality-Scope based evolutionary algorithm would then be recalled to optimise the traffic signal solutions as required by the phase-state diagram of the evolving network.
    Download PDF (640K)
  • Phyrum KOV, Yasuhiro HIROBATA, Yasuyuki NAKAHIRA
    2005 Volume 6 Pages 1410-1425
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    As the actual condition of road network exhibits different combination of traffic modes in the light of normal day-to-day variation, the network travel time is subject to uncertainty and variability. The model proposed in this paper suggests an approach to investigate the effect of variation of mixed traffic on travel time and its uncertainty by using an analytic method. Based on this method, the first and second order moments of link travel time can be evaluated while we capture the fluctuation of traffic flow by each mode. An illustrative numerical example is reported on a simple network from which some useful observations are presented providing us an insight into how fluctuation of demand, vehicle occupancy, link capacity improvement, and vehicle characteristic affect network performance under a prevailing condition.
    Download PDF (700K)
  • Yongtaek LIM, Hyunmyung KIM
    2005 Volume 6 Pages 1426-1438
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    Existing k-shortest path algorithms has some weaknesses such as path similarity among determined paths and network expansion for describing turn prohibitions. Path similarity represents that many of the alternative paths derived from the k-shortest path algorithm are likely to share a lots of links, so they could not represent heterogeneity. The turning restrictions popularly adopted in real road may violate Bellman's principle of optimality in searching shortest path, so network expansion technique is widely used to avoid such difficulty. But, this method needs additional works to expand the network.
    This paper proposes a link-based shortest path algorithm to generate dissimilar paths for the travel information in real road network where exists turn prohibitions. The main merit of proposed model is to provide efficient alternative paths under consideration of overlaps among paths to alleviate the path similarity. Another merit is that it does not require extra nodes and links for expanding the network. Thus it is possible to save the time of network modification and of computer running. The algorithm is tested with some example networks and then will be expanded to a dynamic case.
    Download PDF (811K)
  • Panatda KASIKITWIWAT, Anthony CHEN
    2005 Volume 6 Pages 1439-1454
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    In transportation, capacity has traditionally been measured at individual elements of the network, such as links (rail lines, road segments and waterway etc.) and nodes (terminals and signalized intersection, etc.). These measures do not constitute the transportation network capacity. Recently, Wong and Yang (1997) proposed the concept of network reserve capacity to estimate the transportation network capacity, but the concept is restricted to a common multiplier for all O-D pairs. In this study, the concepts of ultimate capacity and practical capacity (with a non-uniform O-D growth) are applied to the transportation problem to relax the limitation of the reserve capacity concept; thus, the concepts can yield information regarding the spatial distribution of the demand pattern. The definitions of three different network capacity concepts and their models will be described. The analysis of network capacity for three different concepts is provided to demonstrate their features.
    Download PDF (507K)
  • Qiang MENG, Guoshan LIU, Hai YANG
    2005 Volume 6 Pages 1455-1470
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    This paper deals with a class of transportation network optimization problems with user equilibrium constraints, which can be formulated by the bilevel programming models. It first shows that the linear independent constraint qualification at any feasible path flow does hold for the user equilibrium traffic assignment problem with parameters. Accordingly, the fact that the transportation network optimization problems with user equilibrium constraints have a single level continuously differentiable optimization model is verified again. In view of the unique characteristics of a transportation network problem, a novel trust region method for the transportation network optimization problems with user equilibrium constraints is then developed. Compared to the typical trust region methods in operations research, the trust region method solves a series of sub-problems with linear objective functions and the customized trust regions for the link flows. However, it is still a convergent algorithm. Finally, this study employs an origin-destination matrix estimation problem as a numerical example to test the proposed trust region method.
    Download PDF (575K)
TRAFFIC FLOW ANALYSIS
  • Rahman Md. MIZANUR, Fumihiko NAKAMURA
    2005 Volume 6 Pages 1471-1483
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    A model of passing-overtaking maneuvers in terms of total traffic volume and percent of rickshaws is presented in this paper; the model is applicable to undivided urban streets with relatively high proportion of rickshaws. Effects of nonmotorized rickshaws on passing-overtaking are also presented. The results suggest that rickshaws have an adverse effect on the passing-overtaking maneuvers. This paper has also attempted to provide a classification of Level of Service (LOS) for heterogeneous traffic composition in urban areas. In this study LOS are firstly defined in six standard groups, such as levels A, B, C, D, E, and F. The operating characteristics considered to define the level of service are: total number of passings -over takings (no/hr/20m) and average speed of passenger cars (km/hr). Based on the observed data the levels of service are classified into four groups from the road users point of view, namely LOS I (Free-flow condition), LOS II (Partial-constraint flow condition), LOS III (Constraint flow condition), and LOS IV (Congested-flow condition). Level of service classification evolved in this study will help to identify deficiencies of an urban streets system and to plan for alternative improvement measures to attain a desired level of service.
    Download PDF (594K)
  • Pin-Yi TSENG, Feng-Bor LIN, Sheng-Long SHIEH
    2005 Volume 6 Pages 1484-1495
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    Free-flow speed is a parameter that is being used extensively for capacity and level-of- service analysis of various types of highway facilities. As part of an ongoing effort to revise Taiwan Area Highway Capacity Manual, this study collects and analyzes free-flow speed data at the midpoints of seventy-six multilane rural and suburban highway segments in Taiwan. The objectives of this study are to develop models for estimating average free-low speed and for identifying the distribution of individual free-flow speeds. The most important variables that govern free-flow speed are vehicle type, speed limit, and the spacing between signalized intersections. The resulting models facilitate both manual analysis and computer simulation of rural and suburban highways.
    Download PDF (737K)
  • Chu Cong MINH, Kazushi SANO, Shoji MATSUMOTO
    2005 Volume 6 Pages 1496-1508
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    Although much knowledge about traffic characteristics was acknowledged, very little attention has been paid regarding motorcycle. The purpose of this study addresses a comprehensive analysis of motorcycle behavior and operation through videotaping of some roads that have significant motorcycle proportion. Four locations in Hanoi, Vietnam have been found to meet criteria for data collection, including exclusive motorcycle lanes, mixed traffic and undivided roadways. Speed—flow relationships were developed for all locations, in which the adjustment factor for the present of vehicles other than motorcycle was based on motorcycle equivalent unit. Statistical analyses of the empirical data were utilized to demonstrate the characteristics of motorcycle speed, time headway regarding to traffic flow. The present paper provides a basic understanding of characteristics of motorcycle traffic. The finding obtains from this research may be used to develop new procedures for Highway Capacity Manual (HCM), which adapt developing countries as well as provide the data needed to develop a motorcycle simulation model.
    Download PDF (597K)
  • Jumsan KIM, Zunhwan HWANG, Sungmo RHEE
    2005 Volume 6 Pages 1509-1517
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    The motive of this study is derived from the problem recognition that the headway of passing vehicle in signalized intersection cannot be merely determined by departing sequence. Traffic speed and headway data of passing vehicles in signalized intersection have been obtained by using magnetic detector (NC 97) and detecting program, and the data was analyzed. Without special treatment, the model established on passing behavior of vehicles was meaningless by the viewpoint of statistic analysis. So, special treatment such as filtering (upper 85% offensive driver driven vehicle's) and log scaling of data was carried on. With this filtered data, meaningful model (where coefficient of determination is 0.91) was established. This model explained the fact that vehicle headway is affected by not only the speed and headway of previous vehicle, but also the speed of itself in a signalized intersection.
    Download PDF (705K)
  • Lawrence W. LAN, Feng-Yu LIN, April Y. KUO
    2005 Volume 6 Pages 1518-1533
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    This paper aims to develop a parsimony procedure to identify if chaotic phenomena exist in a traffic flow dynamics. A wide-ranging comparison of geometric plots and statistics between the observed one-minute traffic data and their surrogates are carried out. The most crucial indexes are selected to establish the parsimony testing procedure, which are successfully validated by known time series data generators and further applied to identification for chaotic traffic time series. We have strong evidence to conclude that chaotic phenomena, rather than random structures, do exist in the nature of freeway short-term (one-minute) traffic during the morning hours.
    Download PDF (1476K)
  • Chungwon LEE, Jeong Whon YU, Hyeok Ryeol YOON, Keemin SOHN
    2005 Volume 6 Pages 1534-1544
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    This paper presents the two applications of the two-fluid model: characterizing the network performance of a subnetwork in Seoul (“Seoul Kangnam network”), and analyzing the snowing impact on the network wide performance. More than ten thousand probe data each day were collected and analyzed. The data contains moving and stopping times with their positions. According to the comparison with other cities in literature, the Seoul Kangnam network has lower parameter n (about 0.90) and higher parameter Tm (about 2.17) in normal traffic condition. Therefore, the Seoul Kangnam network turned out to need operational improvement during off-peak period but to keep sustainable network performance during peak period. The prominent parameter change was observed. About three days were needed to recover the normal network performance after snowing in the example cases. This paper shows the potential usefulness of the two-fluid model in quickly evaluating various network wide impact including snowing, raining, construction etc.
    Download PDF (725K)
  • Haruka SUZUKI, Prakash RANJITKAR, Takashi NAKATSUJI, Yasushi TAKEICHI
    2005 Volume 6 Pages 1545-1556
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    Some alternative driver models are proposed to mimic accelerating/decelerating behaviors of drivers particularly at curves and intersections. The models are calibrated against experimental data sets taken from car-following experiments conducted in a test track in Hokkaido, Japan. The position and speed data of vehicles were measured precisely using Real Time Kinematic (RTK) GPS receivers. The proposed models are validated against the General Motors (GM) model, a well-known car-following model. The comparison gave a clear advantage to the proposed mixed models over the other alternative models and the GM model, producing the lowest RMSE and highest correlation values.
    Download PDF (542K)
  • Takamasa IRYO, Kohji INOUE, Takahiro TOHYAMA, Yasuo ASAKURA
    2005 Volume 6 Pages 1557-1569
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    This study analyzes time-dependent correlations between travel time and traffic volume in urban expressways. Traffic volume can be changed by many factors and is necessary to know what factors are important. Congestion can be one of the most important factors. It is natural to consider that drivers avoid congested roads and this can make negative correlations between travel time and traffic volume in neighbour hour. In addition, congestion may influence traffic volume in remote hours. It is important to check such “time-dependent” correlations to examine policies like dynamic congestion pricing and so on. This study analyzes the time-dependent correlation between travel time and traffic volume with data obtained with traffic detectors. As a result, negative correlations are observed in neighbour hours and positive correlations are observed between daytime travel time and night traffic volume. This indicates that congestion can have relationship to traffic volume in later hours.
    Download PDF (1610K)
  • Hsun-Jung Cho, George C. P. Tsai
    2005 Volume 6 Pages 1570-1581
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    The purpose of this research, it is analytically the interrelations between the traffic flow characteristics that include traffic density, rate of flow, and speed (k, q, and u) into completely congestion traffic conditions. This study investigates the macroscopic models of traffic flow characteristics on a new weaving section on the Taiwan National Freeway Systems, focusing on the interrelations between the traffic speed, density, and rate of flow of the congestion traffic conditions. The analysis based on these driver behaviors includes 100% lane-changing and 65.22% weaving activities. The results for this study show the fundamental interrelation of traffic flow characteristics on the completely congestion traffic flow condition similar to Greenshield model and showing no influence on lane changing and weaving activities. These regression models concludes that the critical density is 55.74 vehicles per kilometer, the speed is 34.29 kilometer per hour, and the maximum rate of flow (capacity) is 1854 vehicles per hour per lane, respectively. These results represent that the value of traffic speed and rate of flow are same the Highway Capacity Manual (HCM) and the traffic density is more then the HCM.
    Download PDF (513K)
  • Prakash RANJITKAR, Takashi NAKATSUJI, Akira KAWAMUA
    2005 Volume 6 Pages 1582-1596
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    Car-following model has important applications in traffic and safety engineering e.g. for traffic simulation, Advance Vehicle Control and Safety System etc. A great deal of investigation works were conducted in the last five decades to model the longitudinal interaction between adjacent vehicles as a result numerous models are available now. A performance based benchmarking of these models might be useful to evaluate their capabilities in representing real driving behavior. Data precision is the key factor to make any such evaluation meaningful. RTK GPS is the latest technology in data acquisition that makes it possible to acquire high resolution vehicular movement data at an outstanding level of accuracy. Several car-following models were evaluated based on test track experiment data using a GA based optimization method. It was interesting to see a simple linear model performing better than some sophisticated models.
    Download PDF (673K)
  • Juan Paulo J. DE GUZMAN, Alexis M. FILLONE
    2005 Volume 6 Pages 1597-1609
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    Various traffic management schemes have been implemented through the years to solve, if not minimize traffic congestion along EDSA. The most recent one is the strict implementation of the Yellow Lane Policy (YLP). Under the said scheme, buses were segregated into their own lane, separate from other vehicular traffic. This study evaluated the effects of the said policy by examining vehicular flow during the morning, noon and afternoon peak periods conducted before and after the strict implementation of the said policy. The vehicular flow values were then multiplied by the average vehicle occupancy, hence coming up with commuter-flow values. This set of values was then used as the dependent variable in a natural log model regressed against the number of violations as the dependent variable. Ultimately, analysis showed that during the time of the observation, the YLP has not improved commuter flow along the chosen segments of EDSA.
    Download PDF (971K)
  • Feng-Bor LIN, Pin-Yi TSENG
    2005 Volume 6 Pages 1610-1623
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    Current methodologies used in many countries for capacity and level-of-service analysis of signalized intersections are based on the concept of saturation flow. They assume that, after green onset, the discharge rate of queuing vehicles will quickly reach its saturation flow rate after four or five vehicles have entered the intersection, and that the saturation flow will be sustained until shortly after the signal change interval begins. Recent studies, however, show that this saturation flow model often cannot realistically represent the observed queue discharge behaviors. Therefore continued use of the traditional model will have serious implications. This paper uses field data collected in Taiwan and the United States to analyze first the discrepancies between the observed queue discharge behaviors and the saturation flow model. It then discusses the implications of continued use of saturation flow for capacity estimation.
    Download PDF (734K)
  • Feng-Yu LIN, Lawrence W. LAN
    2005 Volume 6 Pages 1624-1636
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    This paper examines the nonlinear dynamics of traffic time series data measured by different time scales, including one-minute, five-minute and ten-minute counts. The approach of Lan's et al (Journal of the Eastern Asia Society for Transportation Studies, Vol. 6, 2005) parsimony procedure is employed to test for chaos. The empirical results from the United States I-35 Freeway field data reveal that if the flow data are measured in one-minute, the traffic dynamics in the morning hours exhibit chaotic phenomena, However, if the same flow data are measured in five-minute and ten-minute, chaotic structures may disappear, and instead, quasi-periodic motions may emerge.
    Download PDF (717K)
TRANSPORTATION DEMAND MODELING
  • Kali Prasad NEPAL, Daisuke FUKUDA, Tetsuo YAI
    2005 Volume 6 Pages 1637-1650
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    The primary purpose of this research is to analyze theoretically the group decision-making process leading to the allocation of activity time and the consumption of goods, with particular emphasis in the households. Each household is characterized as a group of individuals making joint decisions about their activity participations, alternative activity time allocations and consumption of various goods, such as independent and joint activity time allocations as well as private and shared consumption patterns. We firstly explore why individual-based models are not realistic in multi-person households, and secondly, we summarize various intra-household activity time allocation models based on different decision-making processes. All models are presented under microeconomic principle of utility maximization to represent the economic behavior of the households.
    Download PDF (474K)
  • Tsung-Hsien TSAI, Chi-Kang LEE, Chien-Hung WEI
    2005 Volume 6 Pages 1651-1666
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    This paper develops two dynamic neural network structures to forecast short-term railway passenger demand. The first neural network structure follows the idea of autoregressive model in time series forecasting and forms a nonlinear autoregressive model. In addition, two experiments are tested to eliminate redundant inputs and training samples. The second neural network structure extends the first model and integrates internal recurrent to pursue a parsimonious structure. The result of the first model shows the proposed nonlinear autoregressive model can attain promising performance and most cases are fewer than 20% of Mean Absolute Percentage Error. The result of the second model shows the proposed internal recurrent neural network can perform as well as the first model does and keep the model parsimonious. Short-term forecasting is essential for short-term operational planning, such as seat allocation. The proposed network structures can be applied to solve this issue with promising performance and parsimonious structures.
    Download PDF (774K)
  • Yu-Chin CHEN, Liang-Shyong DUANN, Wan-Pei HU
    2005 Volume 6 Pages 1667-1681
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    In this paper we discuss the finite-sample properties of maximum likelihood estimators of multinomial probit model. Sample sizes ranging from 20 to 200 at interval of 10 are tested. For each sample size, we estimated 9 utility specifications under different error covariance matrix assumptions. Robust variance matrix is used to get the asymptotic covariance of estimators for each model. We use Monte Carlo experiments to generate the data sets and replicate 500 times for each experiment. We found that when the size of standardized parameter of generic variable is greater than 0.5 and the sample size is greater than 60, the odds of getting significant estimator that is close to the true value are larger than 0.9. When the size of standardized parameter is small, e.g., less than 0.25, a sample size of at least 170 is needed to get good results.
    Download PDF (900K)
  • Yukimasa MATSUMOTO, Yosuke HORIBA, Motohiro FUJITA, Hiroshi MATSUI
    2005 Volume 6 Pages 1682-1694
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    A model for estimating path flows and modifying O-D flows simultaneously is proposed, in which probe vehicle information and traffic counts are used. The probe vehicle information is accumulated to obtain prior information on path proportions. The link traffic counts are used as sectional volume information of links. The proposed model is constructed as a two stage model. The first stage model is estimation model of path flows and the second stage model is modification model of O-D flows. The proposed model was applied to the Sioux Falls network. As a result, the accuracy of estimated path flows changed in proportion with a ratio of probe vehicles. Moreover, it was clarified that the proposed two stage model had enough ability to estimate path flows and to modify O-D flows accurately even when prior O-D flows had inconsistency with the current traffic conditions.
    Download PDF (575K)
  • Juyoung Kim, Seungjae Lee
    2005 Volume 6 Pages 1695-1707
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    The goal of this paper is to develop the freeway Origin-Destination (OD) demand estimation model using real-time traffic data collected from Freeway Traffic Management System (FTMS). Although it is necessary for the dynamic OD demand between on and offramps to perform more effective traffic management strategies, Automated Vehicle Identification (AVI) systems are unable to assist to collect OD demand due to the limitation of construction and maintenance costs. The existing models use the simulation model to get a link distribution ratio of dynamic traffic flow by time process. It is difficult to load at FTMS and estimate a dynamic OD between on and off-ramps. The formulation of methodology proposed in this paper includes traffic flow techniques and dynamic OD demand estimation techniques using a real-time detector data. The proposed methodology is evaluated by using the real-time data of NAEBU Ring Road, a circulatory freeway system in Seoul, South Korea.
    Download PDF (616K)
  • Wahju HERIJANTO, Neil THORPE
    2005 Volume 6 Pages 1708-1723
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    This research aims to enhance the accuracy of the singly constrained gravity model by incorporating principles from fluid analogy theory. The method is based on an iterative process to generate and allocate trips in an origin-destination matrix and involves estimating the number of trips produced by and attracted to individual zones. The deterrence function and the attraction factor embedded within the model plays an important role in the relative attractiveness of each zone, while the fluid analogy method is used to allocate trips between zones. The method is applied to replicate the distribution of trip-lengths to work by motorcycle in the case-study city of Surabaya, Indonesia for a base year (1993) and for a horizon year (2003). The computed D-statistic and chi-square values indicate that the performance of the combined fluid-analogy and singly constrained gravity model output is better than the conventional singly constrained gravity model in this context
    Download PDF (570K)
  • Yu-Chin CHEN, Liang-Shyong DUANN, Wan-Pei HU
    2005 Volume 6 Pages 1724-1739
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    In this paper we discussed the estimation of choice model with correlation and/or heteroscedasticity across alternatives when the number of alternatives was large. We examined the performance of randomly drawn choice set approach and Poisson regression approach. We designed 6 Monte Carlo experiments. 1000 observations with 30 alternatives each were generated for each experiment. We found that the randomly drawn choice set approach is only applicable to independent and homoscedastic alternatives. When there are correlated factors among alternatives, this approach can not get good fit for the correlated factors even when the number of draw is quite large. We proved theoretically and empirically that Poisson regression model can get the same estimates as some logit type models that showed heteroscedasticity and correlation across alternatives. Poisson regression model can also get estimates close to those of logit kernel model with less computing time.
    Download PDF (605K)
  • Quentin K. WAN, Hong K. LO
    2005 Volume 6 Pages 1740-1750
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    Random utility models are often used to model traveler mode choice behavior in the demand analysis of a transit system. Transit operators might base their policy and operations planning on these estimates. The choice probability of these models is expressed in terms of utility for all transit modes, which is in turn associated with the characteristics or attribute values perceived by each traveler. In practice, however, often observed or published attribute values (e.g., fare) rather than perceived attribute values are used to estimate the utility. In reality, perception biases of the actual attribute values are likely to happen due to lack of convenient access to information. Fare and travel time, for example, are often not precisely known to all travelers. Therefore, there are variations in the perceived transit service attributes among travelers. How these perception variations affect the choice behavior is not generally understood. Will the provision of perfect information, and therefore the removal of these biases in attribute perceptions, be beneficial to travelers, operators and the overall system? This study conducts some simulation experiments to explore these issues.
    Download PDF (637K)
  • Vu Anh TUAN, Tetsuo SHIMIZU
    2005 Volume 6 Pages 1751-1765
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    The rapid increase of motorcycle ownership associated with inadequate public transport has largely contributed to increased traffic congestion, accidents and environmental pollution in Hanoi. A policy framework to improve public transport and control motorcycle ownership simultaneously has become necessary for solving the problems. Therefore, practitioners have shown interest in the household motorcycle ownership behavior and how households respond to transport policies. Objectives of this study are to develop dynamic discrete choice models expressing the ownership behavior and investigate household response to policy. First, a retrospective survey is conducted for collecting information of the household motorcycle transaction processes. Second, heterogeneity is analyzed using random coefficients logit model. State dependence is investigated using buy-smooth and lagged dependent variables. Finally, the results show the increases in number of workers or students, motorcycle price, income, and previous transactions significantly influence current transaction decisions. Sufficient high taxes imposed on motorcycle users could be effective in controlling the ownership.
    Download PDF (479K)
  • Sony Sulaksono WIBOWO, Berlian KUSHARI, Saksith CHALERMPONG, Kasem CHO ...
    2005 Volume 6 Pages 1766-1774
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    In transportation demand modeling, it is a common practice to utilize discrete choice models in estimating people's choice upon several modes of transportation based on observed socio-economic backgrounds and mode-specific features. The nature of the models, however, dictates the requirement of providing a substantial number of observations. An interesting question is whether it would still be possible to estimate a reasonably good model given only a small amount of data set available. One way to make this small number matters is through simulation. This paper reports an exercise of applying bootstrap technique to generate new data sets from a small number of observations and to estimate new models from them. From this simulation, distributions of models' parameters can be provided, and hence the parameters and their standard errors can be calculated from their respective distribution, and their biases can be corrected. The performance of this simulation-based model is then discussed.
    Download PDF (461K)
TRANSPORTATION SURVEY
  • Satoshi TOI, Yoshitaka KAJITA, Tetsunobu YOSHITAKE, Hiroshi TATSUNI
    2005 Volume 6 Pages 1775-1789
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    We have developed a model that enables us to estimate the traffic volume at non-observed road sections. This model needs the observed traffic volumes at other sections and the constitutional ratio of OD volume on each section in the road network. In this paper, the approximate equation of the error that is produced in the process of the estimation is introduced, and then the reliability of the estimation method is analyzed. Finally the effective and essential links for estimation of traffic volume were extracted from the road network including Fukuoka City Region using the branch and bound method.
    Download PDF (405K)
  • Wen HANG, Xu-hong LI, Peng JU, Jie HE
    2005 Volume 6 Pages 1790-1803
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    In 2004, with the increasing requirements of restriction upon overloading in Anhui, a province-wide comprehensive overloading transportation survey has been conducted to evaluate the overloading status quo and enforcement efficiency with the support of the World Bank. A total of six site surveys were conducted at Hefei, Fuyang, Luan, Wuhu, Huainan and Huangshan Areas with four main contents: traffic volume, axle load, freight information and registration information.
    Via statistical analysis on the survey data, conclusions were reached as the following: vehicle overloading is very universal and serious at arterial highways in Anhui nowadays. The traffic loads have far exceeded the designed bearing capacity of highways and have caused prevalent premature pavement damage, especially for rigid pavements. Actually, the overloading trucks are now engaged in highway freight transportation due to the disordered overloading enforcement strategies and the deficient inspecting technologies.
    Download PDF (734K)
BEHAVIORAL ANALYSIS
  • Chieh-Hua WEN, Chen-Lin TSAI
    2005 Volume 6 Pages 1804-1817
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    This paper examines potential travel behavioral changes of in response to the implementation of electronic tolls by distance traveled and time-of-day. An interactive computerized questionnaire with a stated preference experiment was designed to interview auto travelers using freeways for their daily commuting. The joint multinomial logit and nested logit models are employed to identify important factors influencing route and departure time choices. The preferred nested logit model implies a high correlation between unobserved utilities of two route alternatives (freeways and non-toll roads) in the earlier departure time nest and rejects the choice structure in which the choices of route and departure time are jointly determined. Elasticity analyses indicate that freeway travelers are more sensitive to total travel time than toll; existing short-distance toll-free travelers are more sensitive to the toll levels than other travelers and likely to alter their travel choices when implementing electronic tolls.
    Download PDF (666K)
  • Hsin-Li CHANG, Shuen-Cheng WU
    2005 Volume 6 Pages 1818-1832
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    This paper is aimed to explore the factors that might affect the elderly people's mode choice behavior in Taiwan. Under the structure of disaggregated demand analysis, a conceptual framework on elderly people's mode usage is formulated. All kinds of mode usage of elderly people are categorized into three alternatives, which are related to the explanatory variables by a multinomial model. Empirical study is conducted through the data released from the Institute of Transportation of Taiwan. Empirical results show that age, gender, employment status, and living environment are proven to be the significant factors that influence mode usage of the elderly in Taiwan. By overcoming the limitation of data sources, we do conduct a pioneer study in discussing the mode choice of elderly people in Taiwan. Conclusions from this study can serve as a worthy and valuable example for Asian countries that are also confronting the problem of population aging.
    Download PDF (634K)
  • Hye Ran KIM, Kyung Soo CHON
    2005 Volume 6 Pages 1833-1843
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    Real-time traffic information is considered as system which can alleviate traffic congestion. Effective information will help drivers make efficient travel decisions, and improve network performance. To make efficient strategy of providing information, the analyzer should quantify the reduced congestion by the traffic information. Many other route diversion models considered only media-information as information sources and ignored the influence of on-site information such as visible delay or average speed, which in practice affects drivers' route diversion decision. In those models, the influence of on-site information could be disguised to the influence of media information, and the effectiveness of media information would be over-estimated. The proposed model including the influence of on-site information will contribute to estimate accurately the influence of information on drivers' route diversion decision and network performance.
    Download PDF (555K)
  • Gebeyehu MINTESNOT, Shin-ei TAKANO
    2005 Volume 6 Pages 1844-1857
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    In snowy regions, there is always transportation challenges associated with seasonal changes. Until now there is no study conducted on modeling the modal choice variation of different seasons for snowy regions. Therefore this study aims to build high school students mode-choice behavior model to predict the probability of students modal choice under seasonal constraints. From the analysis result, it is observed that there occurs a modal shift from bicycle in the summer to walk (for their short distances trips), bus and passenger car in the winter. Therefore, this study tries to form the relative importance of attributes that affect the high school students mode-shift from bicycle in the summer to other modes in the winter. The result is used to suggest policy and strategic options to facilitate high school students trips.
    Download PDF (841K)
  • Mincheol PARK, Suhyeon KIM, Changho PARK
    2005 Volume 6 Pages 1858-1872
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    The peak hour congestion is derived by concentration of commuters in a big city. To avoid the traffic congestion, commuting drivers change their original departure time or route based on traffic information and previous experiences, and by doing that, commuters can arrive at workplace on time. In this study, the relative importance of factors influencing on commuters' switching behavior was analyzed using binomial probit model, and the frequency of traffic information uses was analyzed using count models such as negative binomial and Poisson regression model for both the home-to-work and work-to-home trip in Seoul. The estimation results showed that there are remarkable differences from commuters' switching behavior of home-to-work and work-to-home trip, and provided the insight into relative effects of factors on the frequency of pre-route and en-route information uses.
    Download PDF (875K)
  • Jin-Long LU, Chih-Wen YANG
    2005 Volume 6 Pages 1873-1887
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    This paper explored the impacts of incomplete travel information on airline passengers' choice behavior in practically. We specified three different kinds of utility function to incorporate the effects of incomplete information into choice model and used binary logit model and random coefficient logit model (RCL) to estimate, based on the data collected from a stated choice experiment. From the estimation results, it can be found that the attributes with incomplete information, included airline safety and Consumers' word-of-mouth, were less important than the variables with complete information, such as flight frequency as well as ticket price, after considering the effect of the taste variants between individuals. Further, it was also found that passengers' backgrounds interacted with the incomplete information attributes would significantly have negative impacts on their choices. The higher the passengers' income as well as education, the more negative impacts of incomplete information on their choice behavior.
    Download PDF (595K)
  • Kyunguk CHANG, Kyungsoo CHON
    2005 Volume 6 Pages 1888-1903
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    In recent years, the importance of Car-following models has increased significantly, although car-following behavior has been studied for an half of a century. In most cases, the simulation model results are used in actual field. However, correct verification about their result is not attained.
    Most models do not consider about reaction time as various and in reality, it is general to use fixed value or random variable for the reaction time. As a result, it causes the rigidity of models, and the “time difference” between the actual behavior and models so, acceleration and deceleration at time (t) are simultaneously. Therefore, reaction time is used as lognormal distribution, and “perceptual threshold” was estimated from the acceleration rate versus relative speed to improve acceleration and deceleration rate.
    Goodness of fit of the model is examined by Root Mean Squared error (RMSE) comparing to other models. Consequently, the new model performed better description of actual driver's behavior.
    Download PDF (548K)
  • Tsan-Huang HUANG
    2005 Volume 6 Pages 1904-1913
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    It is important for the design of a reasonable traffic signal to fit the level of services of intersection and traffic characteristics relating to driver's psychological response that arises depending the waiting time. In this paper, a method for estimating and/or predicting the driver's psychological response to the condition when seeing a rough traffic flow is proposed based on fuzzy set theory. Concretely, when drivers waiting to a traffic signal are exposed to the impact of psychological factors, the fuzzy relation between level of services of intersection and traffic flow characteristics of the random image are approximately given a bi-variate membership function. Next, a method for evaluating the driver's psychological response is proposed by introducing the concept of fuzzy probability. Finally, the validity and applicability of the proposed method are confirmed experimentally by camera data. The experimental results are in good agreement with the theory.
    Download PDF (768K)
  • Chieh-Hua WEN, Ming-Jyh WANG, Lawrence W. LAN
    2005 Volume 6 Pages 1914-1928
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    This paper develops a multinomial logit (MNL) model to identify important factors affecting the selection of automobile insurance policies (AIPs), which are characterized with high similarity among the bundled alternatives. Various nested logit (NL) model specifications are attempted to elucidate the choice behaviors for these highly similar auto insurance products. Some 3,000 voluntary automobile insurance records, randomly drawn from a non-life insurance company in Taiwan, are used to conduct the empirical study. The estimation result of the preferred MNL model shows that vehicle usage, engine capacity, imported/domestic vehicle, and number of claims are significant factors influencing one's selection of AIPs. Owners of newer vehicles, of larger engine capacities, of imported vehicles, and of more insurance claims tend to purchase wider coverage of insurance packages. The NL models confirm that the bundled AIPs are similar and highly correlated products. The empirical results also provide evidence of adverse selection in automobile insurance market -- the wider the insurance coverage, the less motivation the insured would reduce the number of claims (or equivalently, prevent the accidents).
    Download PDF (1201K)
  • Aphisit PHOOWARAWUTHIPANICH, Tuenjai FUKUDA, Sorawit NARUPITI, Atsushi ...
    2005 Volume 6 Pages 1929-1942
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    A study of the potential use of a new transportation mode often challenges transport planners in forecasting future demands. Various factors influence travelers on making decision, especially attitudes, demography, and travel characteristics leading to, if not properly examined, unreliable demand forecasting. A Car-sharing system is an innovative transportation mode for Bangkok people to shift from an existing mode. This paper focuses on a study of the travelers' opinion on the new mode and the potential use of the Car-sharing system in Bangkok regarding pre- and post- attitudes toward the system. A two-phase analysis of a preliminary study and an attitude-related research are conducted to elicit travel behavior information associated with an introduction of the Car-sharing system. Descriptive and conductive statistics are presented for describing the relationship between the level of interests and the travelers' explanatory characteristics. The study develops a methodology to treat the attitude variables by group classification and then incorporate attitudes in modeling process (Binary Logit Model). In addition, two classification methods (Statistical and Judgmental) are considered and compared. The study brings about the understanding on the effect of attitudes toward the Car-sharing system resulting in more accurate estimation of the use of the system.
    Download PDF (529K)
  • Kazutaka TAKAO, Yasuo ASAKURA
    2005 Volume 6 Pages 1943-1955
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    We are studying to model route choice behaviour with words because travelers' psychological states are observed directly and clearly by words from a bottom-up viewpoint. Route choice behaviour is expressed in two stages: “recognize” and “choose” stages. Their thinking processes are clearly described by placing “cognition results” between the two stages. Cognition results are their impressions and feelings of each travel route. This paper shows a method that extracts cognition results out of open-ended questionnaire texts. Once we collect the linguistic knowledge, cognition results can be extracted from texts even if they are unseen. Not only simply extracting words that express cognition results, the extraction of denial is also reported to interpret properly as cognition results. Finally, an experiment on the extraction of cognition results from unseen texts is reported.
    Download PDF (569K)
  • Yoriyasu SUGIE, Junyi ZHANG, Akimasa FUJIWARA, Koji SUTO
    2005 Volume 6 Pages 1956-1970
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    In this paper, time choice models to and from work are developed using panel data collected during a three-year period after the introduction of a flextime system in a company. Interesting facts revealed: (1) the habitual effect of arrival time to work continues after the new system is implemented, but decreases over time, (2) household structure does not have a great effect on arrival and departure times to and from work, and (3) interdependence between the two times to and from work are significant. In light of these conditions, it is believed that the times to and from work are decided simultaneously, hence a simultaneous choice model is formulated and estimated based on the same data. The interdependence between these two times is significantly confirmed in the model.
    Download PDF (519K)
  • Tuenjai FUKUDA, Shigeru KASHIMA, Atsushi FUKUDA, Sorawit NARUPITI
    2005 Volume 6 Pages 1971-1986
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    This paper analyzes the possibility of car sharing application in Bangkok by focusing on the potential target users regardless of socio-economic characteristics and modal choice behavior with respect to their perception and responsiveness on the assumed car sharing's service attributes, particularly what factors affect their decision on making choice probability. Demand forecasting on the use of car sharing system is performed. Introducing car sharing, as being well-known system in Europe and North America as an innovative mobility for substituting vehicle ownership, seems to be an appropriate remedial countermeasure to tackle the rampant increase of private vehicles because car sharing allows different users to share use a fleet of diversified vehicles several times a day with convenient and privately environment. The findings revealed that factors affecting a user's decision making on choice probability are socio-economic like income, occupation, and education as well as service attributes like travel time and costs.
    Download PDF (625K)
TRANSPORTATION DEMAND ANALYSIS
  • Iv LIM, Yasuhiro HIROBATA
    2005 Volume 6 Pages 1987-2002
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    This study aims firstly at examining the actual condition of travel behavior changes on both weekdays and weekend days in a local city of Japan by using a series of data from person-trip surveys conducted in 1977, 1991 and 2001. Secondly, it is to show that these travel behavior changes were determined by their factor attribute changes and their factor composition changes. Thereby, quantitative understanding of these temporal macro changes is attained, and it can find out direction of appropriate analysis required for future travel demand prediction. Consequently, the recent change of individuals' structures does not have as much effect on trip generation as the individuals themselves who reduce their trip generations. Dramatic modal split change such as car shift was caused by the change of traveler's mode itself rather than by the change of OD trip patterns, and trip attraction change was caused by the change of destination choice from other zones rather than that zone changed its trip generation.
    Download PDF (706K)
  • Chaug-Ing Hsu, Yu-Chiang Chen, Hui-Chieh Li
    2005 Volume 6 Pages 2003-2018
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    This study incorporates intercity travelers' mode choice as well as route choices into a framework by exploring the relationship among key demand and supply variables. Travelers' optimal mode and route choices along a corridor with air transportation (AT), high-speed rail (HSR) and automobiles are assumed by minimizing their generalized travel time including travel time and cost time, which is transferred from travel monetary cost via the value of time. Travel time is defined as including waiting time at stations, riding time, access time and egress time, while travel monetary cost is composed of fare and access/egress cost. The optimal alternative is explored by comparing, respectively, differences in travel time and in cost time between two alternatives. A case study is presented to demonstrate the application of the model. The results show that travelers with the higher values of time tend to choose faster modes as their intercity modes, which are combined with progressive or backtracking taxis to the final destinations. The results also show regardless of trip distance, the tendency towards using automobiles to complete the entire trip is higher for those with the lower values of time.
    Download PDF (547K)
  • Yoshitaka KAJITA, Satoshi TOI, Yuho KUBO, Hiroshi TATSUMI
    2005 Volume 6 Pages 2019-2030
    Published: 2005
    Released on J-STAGE: March 14, 2008
    JOURNAL FREE ACCESS
    In this paper, the structure of a time-of-day travel demand forecasting model system is discussed. A short-time Origin-Destination [OD] model is constructed to estimate traffic volume in each time class during a day. A short-time traffic assignment is also proposed considering some trips do not always arrive at a destination within the same time class and link into the next phase. Also, Sequential short-time modal split must consider traffic conditions of each time class. Therefore, joint modal split and assignment model is constructed. In these models, time distributions and trip-length are required. These models are analyzed considering differences in characteristics of trip purposes and personal attributes and estimated by model parameters from the Person trip survey data.
    Download PDF (775K)
feedback
Top