Bongsoo Son

Analysis of traffic accident size for Korean highway using structural equation models

Accident size can be expressed as the number of involved vehicles, the number of damaged vehicles, the number of deaths and/or the number of injured. Accident size is the one of the important indices to measure the level of safety of transportation facilities. Factors such as road geometric condition, driver characteristic and vehicle type may be related to traffic accident size. However, all these factors interact in complicate ways so that the interrelationships among the variables are not easily identified. A structural equation model is adopted to capture the complex relationships among variables because the model can handle complex relationships among endogenous and exogenous variables simultaneously and furthermore it can include latent variables in the model. In this study, we use 2649 accident data occurred on highways in Korea and estimate relationship among exogenous factors and traffic accident size. The model suggests that road factors, driver factors and environment factors are strongly related to the accident size.

Estimation of the optimal number of cluster-heads in sensor network

A sensor network system consisting of a large number of small sensors with low-power can be an effective tool for collection and integration of data by each sensor in a variety of environments. The collected data by each sensor node is communicated through the network to a single base station that uses all collected data to determine properties of the data. Clustering sensors into groups, yields that sensors communicate information only to cluster heads and then the cluster-heads communicate the aggregated information to the base station. We estimate the optimal number of cluster-heads among randomized sensors in a bounded region. We derive solutions for the values of parameters of our algorithm that minimize the total energy spent in the wireless sensor network when all sensors communicate data from the cluster-heads to the base station. Computer simulation shows that the energy consumption reduce as the optimal number of cluster-heads for the proposed method.

Shock Wave and Cumulative Arrival and Departure Models: Partners Without Conflict

The shock wave and cumulative arrival and departure curve techniques are staples of traffic analysis, particularly for the modeling of freeway congestion, but the literature contains numerous suggestions that they are not compatible. Three numerical examples illustrate that although the two models define queue lengths in different ways, they are compatible, yield identical estimates of travel time and delay, and provide additional insight into the nature of freeway congestion when used together instead of individually.

Bus Arrival Time Prediction Method for ITS Application

It is known that stoppage times at signalized intersections cause the biggest errors in bus arrival time prediction for real-time Bus Information System (BIS) services and no particular method is proven successful so far. This study developed a prediction method that compares the predicted bus travel times from bus stop to the stop line at signalized intersections by using Kalman filtering technique with the state of green time indications of traffic signals, and then incorporates possible stoppage into a next link travel times. From field surveys and in-lab simulation, the proposed method was found superior to other conventional techniques showing an average of more than 200% improvement in prediction accuracy.

Development of Real-Time Optimal Bus Scheduling and Headway Control Models

Infrequent bus service because of a sudden rise in international gas prices has created a significant social problem in South Korea. Bus service frequency is determined by considering bus operation costs to the service supplier, passenger demand, and so forth. Optimal bus frequency was determined and a timetable for each bus stop was created by applying the DTR (demand- and travel time–responsive) model or the DTRC (demand- and travel time–responsive model for critical scheduling areas); both use data for each bus stop and route segment. A bus headway control model that can monitor and assess punctuality of an operation was also developed. This model was based on real-time event data, such as bus stop departures and arrivals for buses operating on a line-based timetable, constructed for each bus stop. It was also based on the bus stop traffic cost, which was applied in the bus scheduling model. The models ability to evaluate the punctuality of operation made it possible to transmit headway control instructions (when to decelerate or accelerate) to a bus driver via an onboard unit. A model verification process was implemented by using data collected from a bus management system and integrated transit fare card system for the bus route in Seoul. To evaluate the reliability and uncertainty of the optimal solution, a sensitivity analysis was implemented for the various parameters and assumptions used in the models.

Traffic Flow Forecasting Based on Pattern Recognition to Overcome Memoryless Property

A variety of methods and techniques have been developed to forecast traffic flow. Current nearest neighbor non-parametric traffic flow forecasting models treat the dynamic evolution of traffic flows at a given state as a memoryless process; the current state of traffic flow entirely determines the future state of traffic flow, with no dependence on the past sequences of traffic flow patterns that produced the current state. Since traffic flow is not completely random in nature, there should be some patterns in which the past traffic flow repeats itself. In this paper, we proposed a pattern recognition technique, which enables us to consider the past sequences of traffic flow patterns to predict the future state. It was found that the pattern recognition model is capable of predicting the future state of traffic flow reasonably well compared with the k-nearest neighbor non-parametric regression model.

Heuristic algorithm for traffic condition classification with loop detector data

This paper presents a heuristic algorithm for detecting traffic conditions from loop detector information at signalized intersections. With use of the characteristic of occupancy data from each cycle, the algorithm determines the level of traffic conditions. Several variables were introduced for this algorithm, and the detailed descriptions of flow chart of the initial algorithm were included in this paper. The proposed algorithm has a simple logic, however, the performance of the algorithm could be improved by performing calibration or by introducing additional variables.

Probabilistic Model of Traffic Breakdown with Random Propagation of Disturbance for ITS Application

In this paper, a probabilistic model of vehicular traffic breakdown applicable to Intelligent Transportation Systems (ITS) is presented. When traffic demand exceeds freeway capacity, the so-called breakdown occurs and the congestion begins. While preventing the breakdown is a major concern of traffic operation, the mechanism of breakdown is not thoroughly explained and most of the research regarding traffic breakdown rely on empirical analysis. To further our understanding of traffic breakdown, this paper explains the phenomenon of traffic breakdown in terms of random propagation of traffic disturbance and proposes a probabilistic model of breakdown. A Monte-Carlo simulation is also conducted to investigate the characteristics of the proposed model.

A simulation model of congested traffic in the waiting line

Congested traffic shows very complicated and stochastic features. One of the most interesting features is the amplification/decay of perturbations, indicating the development of small speed oscillation in the downstream into the large speed oscillation in the upstream. While traffic theories attempt to explain this phenomenon, the mechanism of the speed oscillation is not yet clear in the aspect of drivers’ behavior. Similar phenomenon is also found in a long waiting line such as in front of the stadium or theater ticket box. In this paper, the stop-and-go movement in the waiting line, which is relatively easy to understand than the vehicular movement, is modeled and simulated. From the simulation, it is found that the amplification/decay of perturbation exists in the waiting line and shows similar pattern as in the vehicular movement.

Variable Speed Limits Model to Minimize Confliction without Change in Total Travel Time Using Queuing Theory

Variable Speed Limits can affect safety and congestion. To estimate the effects of variable speed limits, the operation value should be calculated with queuing theory. In this model after ts minutes since congestion occurred, the queue length is checked and variable speed limits operation would be started. Under this condition, cars pass the modeling area without congestion within the same travel time with non-controlled condition. The confliction under VSL and that under non-controlled condition were compared using SSAM.