Yasuhiro Shiomi

Analysis of Car-following Behavior on Sag and Curve Sections at Intercity Expressways with Driving Simulator

We analyzed the influence of road alignments such as sags and curves and the leading vehicle’s behavior on car-following behavior in a driving simulator experiment. Parameters of a car-following model were estimated from following-vehicle trajectory data collected for 37 participants. Then, relationships between the parameters and environmental factors were analyzed. The results showed that the parameters of the following-behavior model were significantly influenced by expressway alignments such as sags and curves, whereas differences in the leading vehicle’s behavior did not significantly affect the estimated parameters. These findings indicate that measures to assist the following vehicle’s acceleration and deceleration such as adaptive cruise control could be effective in preventing the breakdown of traffic flow.

Experimental Study of Effects of Travel Time Distribution Information on Dynamic Route Choice Behavior

This study focuses on the possibility that providing travelers with information on travel time distribution, along with the existing travel time information, might affect their decision making and enhance traffic control. As an initial step to confirm the effectiveness of travel time distribution information, we use a laboratory experiment to obtain panel data of route choices when information is provided. The rationality of the behavior of the respondents and the characteristics of the route choice are first analyzed by aggregation analysis and then statistically examined by specifying a mixed logit route choice model. As a result, it is revealed that both travel time information and maximum travel time information significantly affect route choice behavior when a penalty is imposed for late arrival.

Network Evaluation Based on Connectivity Reliability and Accessibility

Our daily lives and activities rely heavily on transportation systems, particularly road networks. At the same time, we face the risk of various types of natural disasters. It is, therefore, extremely important to establish road networks that are robust and reliable under conditions of network degradation caused by natural disasters. Even immediately after a disaster, an area should retain connectivity with its neighboring areas, while at the same time residents in the area should have access to adequate urban services, such as medical facilities, at a reasonable travel cost. To achieve the former goal, the concept of connectivity reliability has been developed, and to achieve the latter, accessibility measures have been proposed. This study proposes a method of evaluating road network robustness against a natural disaster by applying the concepts of connectivity reliability and accessibility. A new network evaluation measure, which considers both the risk of link disruption and the number of opportunities in an area, was first established. Then, the measure was applied to test networks, and the influence of network topology and distribution of opportunities were analyzed. Finally, the road network and medical facilities in Kyoto Prefecture were evaluated using the proposed measure.

Model for Location of Medical Facility and Evaluation of Vulnerability and Accessibility of Road Network

This study proposes a method of evaluating the robustness of a road network against a natural disaster and aims to determine the optimal allocation of medical facilities if road network degradation occurs because of severe natural disasters. Connectivity vulnerability and accessibility are important concepts in evaluating the strength and reliability of road networks. This paper first introduces a new network evaluation measure, the distinct paths–based accessibility index (DPAI), which is based on the concepts of both connectivity vulnerability and accessibility; considers the number of distinct paths between a certain origin–destination pair, free-flow travel time, and the number of opportunities in an area; and develops a solution algorithm. A medical facility location model that optimizes the allocation of a medical facility to each centroid according to the DPAI is then formulated, and the results are applied to a real network. Although the optimal allocation of a medical facility varies according to the minimum DPAI, the robustness of an actual network may be improved by reallocating medical facilities. Moreover, the allocation results depend on the value of the importance placed on the number of distinct paths. These results imply that it is critical to set the parameter on the importance of the number of distinct paths by considering risks and geographical features in the subject area.

A trip-chain-based combined mode and route-choice network equilibrium model

ABSTRACT This paper proposes a trip-chain-based mode and route-choice network equilibrium model that considers the ‘common lines problem’ in transit assignment. The proposed model assumes the nested-logit type of mode and route-choice structure for the travellers’ mode and route choice. The proposed model is formulated using the variational inequality problem, and the relaxation method is applied to solve the proposed model. Finally, the proposed model is applied to hypothetical networks, and confirms that the differences in the total mode share, link flows, and link travel times between the proposed model and the traditional trip-based assignment model are not large. We also confirmed that the some trip-chains differed considerably in mode share.

Special issue on measuring, analysing and evaluating transport network reliability

This special issue originated with the organisation of the 6th International Symposium on Transportation Network Reliability (INSTR), which was held in August 2015 in Nara, Japan. The INSTR series is the premier gathering of theworld’s leading researchers and professionals interested in transportation network reliability, where they discuss both recent research and future directions in this increasingly important field of research. This special issue includes eight significant papers on the state of the art in research dealing with measuring, analysing, evaluating andmanaging reliability for transport network systems. Although most of these papers are drawn from those presented at the INSTR 2015 symposium, an open call for papers was alsomade and attracted several submissions. All submitted papers were rigorously reviewed to ensure the quality and correctness of their content. In the first paper, Chen et al. propose an analytical model to determine the optimal cordon toll pricing scheme in a linear monocentric city with a competitive railway/highway system.Assuming that thedaily commuting timeby railmode is deterministic,while that by auto is stochastic and location-dependent across the city, amulti-modal urban systemequilibrium model is developed in which the effects of travel time uncertainty on commuters’ mode choices and households’ residential location choices are considered. The impacts of the level of auto travel time uncertainty on the urban system performances are analytically explored together with that on the design of the optimal cordon toll pricing scheme in terms of toll level and cordon locations. The second paper, by Chen et al., aims to optimise the road pricing scheme tomaximise the reliability of network performance in terms of the network-level travel time coefficient of variation using an efficient simulation-based optimisation (SBO) approach. The proposed approach is demonstrated in a large-scale real-world regional network model by DynusT (dynamic urban system in transportation) to evaluate system reliability as the objective function corresponding to different toll charges for a new toll road in Maryland. The results verify the feasibility of the applications of SBO tomaximise network-wide travel time reliability improvement by solving a dynamic congestion pricing problem. The third paper in this special issue, by Xu et al., examines the impacts of network-wide speed limits. This paper proposes a multi-dimensional modelling approach: (1) a truncated link travel time probability distribution is proposed, and its moments expression is derived to characterise the effect of speed limits on link travel time uncertainty profile, and (2) the risk measure of mean-excess travel time (METT) is adopted as a risk-averse route choice criterion with a complete consideration of both on-time arrival reliability and late arrival unreliability. The paper concludes that a speed limit scheme imposes different impacts on different dimensions of the travel time uncertainty profile, including the statistical characteristics and risk performance.

Assessing Safety of Signalized Intersections

Traffic accidents constitute a serious social problem. Thus, it is essential to identify and remove the risk factors affecting traffic accidents to enhance traffic safety at intersections. This study identified and quantified the main factors influencing traffic accident risk at signalized intersections to propose effective countermeasures. Google Earth was used to collect numerical data related to the geometric attributes of intersections in three regions in Japan: Kagawa, Shiga, and Aichi. A lognormal hurdle model that considered regionality and geometric attributes was then used to quantify factors influencing the risk of traffic accidents involving various types of collisions. The results indicate the existence of significant regional differences in the geometric attributes of intersections in those regions. The findings reveal that intersection size (indicated by the distance between stop lines), length of crosswalks, and setback distance of crosswalks generally and significantly influence all collision risk types; in addition, a compact intersection lowered the risk of all collision types. And in most cases, regional dummy variables were statistically significant. This finding suggests that in addition to intersection geometric attributes, regional differences in some factors influenced collision risk. Thus, it could be reasonably assumed that regionality stemmed from differences in driving characteristics.

Measurement of Travel Fatigue: Objective Monitoring and Subjective Estimation

Using both objective monitoring and subjective estimation, this paper investigates travel fatigue. Objective monitoring included an experiment designed to measure energy expenditure rates with a heart rate monitor while travelers walked on level ground, went up and down stairs, and rode on a bus and subway. Free modulus magnitude estimation was used to investigate travelers’ subjective perceptions of fatigue while traveling. The results of the two parts were compared to understand their relationship. The results showed that the energy expenditure rate was sensitive not only to travel mode, represented by running vehicle acceleration, but also to posture, carrying personal belongings, and in-vehicle crowding. According to participants’ subjective perceptions, the effect of in-vehicle crowding was greater than carrying a 3.8-kg burden; the fatigue rates of standing, both with and without a load, tended to be subjectively overestimated compared with normally transformed energy expenditure rates. Numerical results from a case study in Zhongshan, China, and an online survey of bigger cities in China showed that travel time is not the only issue. Travel fatigue must also be considered in planning urban transportation systems and in evaluating transit service quality and has great potential for predicting choice of travel mode.