Showing papers in "Transportation Research Part C-emerging Technologies in 1994"

An evaluation tool for advanced traffic information and management systems in urban networks

Abstract: Tools for evaluating traffic networks under information supply are a crucial necessity in view of the ATMS/ATIS systems being proposed and implemented around the world as part of Intelligent Vehicle-Highway Systems of the future. This paper presents an evaluation model that incorporates the driver response to information, the traffic flow behavior, and the resulting changes in the characteristics of network paths, into an integrated simulation framework. The model is based on simulating individual vehicle movements according to macroscopic flow principles, the driver path selection behavior under information being explicitly modelled. Detailed modelling of intersection delays as well as a variety of traffic control options for both freeways and arterials are performed. The path-processing component is designed for efficient application of the framework to large and realistic networks. The model can be effectively used for studying alternative information supply and traffic control strategies under various levels of market penetration of in-vehicle ATIS hardware. The paper also discusses its application to candidate networks.

Statistical analysis of commuters' route, mode, and departure time flexibility

Abstract: This paper studies the effects of traffic information on Seattle-area commuters. Models of commuters' route-change frequency, the duration of traffic delay needed to induce a route change, and the influence of pre-trip traffic information on departure-time choice, mode choice, and route choice are estimated. Two modeling techniques are used in these estimations: an ordered logit probability approach and a Weibull duration model with a heterogeneity correction term. The findings of these model estimations provide important insights into how traveler information systems should be designed, implemented, and marketed.

A driving simulator and its application for modeling route choice in the presence of information

Abstract: Models for route choice in the presence of information and motorist reaction to route guidance are currently under development. A major difficulty in developing such models is the lack of appropriate data for testing and calibration. This paper describes a PC-based driving simulator that can be used for collecting relevant data in a controlled environment. The simulator uses 2-D graphics, and consists of three main modules: network performance, guidance generation, and user interface. A flexible design permits the simulation of a wide variety of information systems on any network. The functionality of the driving simulator is demonstrated in a case study with data collected from a group of 10 subjects. The data was used to calibrate a new class of route choice models in the presence of information, which are based on concepts from fuzzy sets and approximate reasoning. The results indicate that until data collected on actual route choice behavior in the presence of information becomes available, appropriately designed driving simulators can become useful tools.

Roadside intelligence for flow control in an intelligent vehicle and highway system

Abstract: The paper reports a design of the flow control function of the highly automated Intelligent Vehicle Highway System that we call SmartIVHS. Work on the architecture, design, experiments, and performance evaluation of this system has been going on for several years. SmartIVHS achieves high throughput and safety through a three-layer control hierarchy distributed between vehicles and infrastructure. Previous work was devoted to the two lowest layers: automatic control of individual vehicles based on on-board sensor information, and coordination of maneuvers by neighboring vehicles. This paper considers the third or ‘link layer,’ which controls the vehicle stream based on aggregate traffic variables. The link layer controller is implemented by roadside computers. The two lowest layers are implemented by vehicle computers. The paper is divided into three parts. First, a structure of the link layer controller is proposed. Its objectives are to maximize throughput and to maintain smooth traffic flow despite disturbances, including lane-blocking incidents. The objectives are met by proper guidance of the speed and lane-changing behavior of vehicles. Second, a macroscopic flow model of SmartIVHS traffic is proposed. A novel feature is the explicit incorporation of the effects of lane changes, entrances, and exits. The program SmartLink simulates this model. Third, performance of the link layer controller is evaluated using SmartLink. The results suggest that the controller maximizes throughput and minimizes delay in the presence of disturbances and incidents.

In-laboratory experiments to investigate driver behavior under advanced traveler information systems

Abstract: In-laboratory experimentation with interactive microcomputer simulation is a useful tool for studying the dynamics of driver behavior in response to advance traveler information systems. Limited real-world implementation of these information systems has made it difficult to observe and study how drivers seek, acquire, process, and respond to real-time information. This paper describes the design and preliminary testing of an interactive microcomputer-based animated simulator, developed at the University of California, Irvine, to model pre-trip and enroute driver travel choices in the presence of advanced traveler information systems. The advantages of this simulator are realized in its versatility to model driver decision processing while presenting a realistic representation of the travel choice domain. Results from a case study revealed that increased driver familiarity with travel conditions and network layout reduced driver reliance on information systems and influences diversion behavior.

Modelling dual carriageway lane changing using neural networks

Abstract: Neural networks offer a potential alternative method of modelling driver behaviour within road traffic systems. This paper explores the application of neural networks to modelling the lane-changing decisions of drivers on dual carriageways. Two approaches are considered. The first, preliminary approach uses a prediction type of neural network with a single hidden layer and the back propagation learning algorithm to model the behaviour of an individual driver. A series of consecutive time-scan traffic patterns, which describe the driver's environment and changes over time as the selected vehicle travels along a link, are input to the neural network, which then predicts the new lane and position of the vehicle. Training data are collected from a human subject using an interactive driving simulation. The trained neural network successfully exhibited the rudiments of driving behaviour in terms of lane and speed changes. A major disadvantage of this approach was the difficulty in recording real-life data, which are required to train the neural network, for individual drivers. The second approach concentrates specifically on lane changing and makes use of a learning vector quantization classification type of neural network. Input to the neural network still consists primarily of time-scan traffic patterns, but the format is changed to facilitate the possibility of data acquisition using image processing. The neural network output classifies the input data by determining the new lane for the vehicle concerned. Performance in both testing and training was very good for data generated by the rule-based driver-decision model of a microscopic simulation. Performance in testing was less satisfactory for data taken directly from a road and highlighted the need for extensive data sets for successful training.

A multiple-path routing strategy for vehicle route guidance systems

Abstract: This paper analyzes the transportation effectiveness of a multiple-path routing strategy using traffic simulation from the perspective of planning and designing a vehicle route guidance system. The test results indicate that the multiple-path routing strategy performs better than the commonly used shortest-path routing strategy.

A planning methodology for intelligent urban transportation systems

Abstract: Recent developments in intelligent transportation systems pose new challenges and opportunities for urban transportation planning. To meet these challenges and to exploit these opportunities, a framework for a new transportation planning methodology has been developed. The methodology operates in a computer environment, called PLANiTS (Planning and Analysis Integration for Intelligent Transportation Systems), designed to facilitate the entire planning process form problem identification, through idea generation and analysis, on to prioritization and programming. To assist in problem identification, PLANiTS provides graphic representation of current conditions, including traffic, air pollution, accidents, and projections of future conditions. A computerized knowledge base, containing information about possible strategies and their effects, and a model base, containing transportation and other analysis models, are used to guide the user in identifying potentially effective strategies and performing the appropriate analysis. To facilitate the use of these tools, PLANiTS provides computer support of group processes such as brainstorming, deliberation, and consensus seeking. PLANiTS is designed for use in urban transportation planning at the local, regional, and state levels; it is intended to support a variety of participants in the planning process including transportation professionals, decision makers in transportation agencies (often local elected officials), citizens, and interest groups. Recognizing that transportation planning is essentially a deliberative, political process, PLANiTS is designed to inform and facilitate, but not replace, the political decision-making process.

Simulation study of driving performance with selected route guidance systems

Abstract: Experiments were conducted in a driving simulator developed by the Hughes Aircraft Corporation to study the human factors aspects of route guidance systems. The primary objective of this research was to study how in-vehicle route guidance system attributes, driver characteristics, and traffic conditions affect driving performance. Four types of route guidance systems were tested. They are: (1) Paper Map, (2) Heads Down Electronic Map, (3) Heads Up Display (HUD) in combination with Electronic Map, and (4) Voice Guidance in combination with Electronic Map. Data were collected for a total of 18 subjects, 9 male and 9 female. All subjects were tested in all four route guidance systems. The following performance measures were collected in the simulator: Number of Navigation Errors and Reaction times to external events. Apart from these, an unobstrusive eye tracker was also used to monitor eye fixations. Data were also collected on driver preferences and subjective workload associated with each of the four route guidance systems. The results of the study can be summarized as follows: (a) Subjective workload, user perceptions, eye tracker dwelling times, and number of errors all indicated that the voice guidance/electronic map combination performed the best, and the paper map the worst. The electronic map was found to be the second best, closely followed by the HUD electronic map. (b) The reaction time modelling yielded slightly different device performance depending on the event being reacted to. The heads up display/electronic map combination performed much better in comparison to its performance in the other performance measures, with voice/electronic map also doing well. The paper map again consistently performed the worst. (c) Driving performance did vary with gender and experience. Not surprisingly, drivers with higher experience performed better than drivers with lower experience. This effect was more prominent among females than males.

Computational experiments with a system optimal dynamic traffic assignment model

Abstract: Although several optimization models have been proposed for the dynamic traffic assignment problem, little computational experience is available. In this paper, we take a first step towards filling this gap by formulating and solving a class of discrete time, nonlinear models that are extensions of those in Carey (1987). An exit function is proposed, and its effects on traffic flow are analyzed using small traffic networks. The performance of three widely available optimizers in solving these models is described. Solution features of both nonlinear and piecewise linear versions of the model are presented and compared with simulation results generated using the Dynasmart simulator.

Dynamic route guidance with queue- and flow-dependent travel time

Abstract: This paper presents an optimal guidance algorithm that takes into account the driver compliance to route advice and flows higher than capacities of links The optimization problem consists in minimizing the travel time of guided vehicles using a model that describes the traffic by a set of flows on a graph The travel time of a link is obtained by computing the delay of vehicles due to queues It leads to Wardrop's travel time when the initial queue is empty and the flow is smaller than the capacity of the link The optimization problem is solved using the Simplex Algorithm recursively on two examples Results show that for most of the demands, the guidance benefit increases with the ratio of equipped vehicles

The role of teamwork in a planning methodology for intelligent urban transportation systems

Abstract: In this manuscript we propose a framework for a transportation planning methodology that recognize the key role that teamwork plays in the decision-making process. Recognizing that the transportation planning process has evolved for many reasons, we develop a cohesive framework for providing intelligent decision support to teams deliberating planning problems. The design methodology considers both user and functional issues in building a matrix of building block functions (BBFs) to support a particular planning process. We illustrate the design methodology by using as an example a recent planning problem from California.

Improved methods for classification of pavement distress images

Abstract: Methods for automated classification of pavement distress images are examined and compared. Images are divided into regions, and a two-stage classification procedure takes place. First, the regions are classified into primitives (plain, longitudinal, transverse, diagonal, or joint), which are the building blocks characterizing the various distress classes. The regional classification results are aggregated and used as input for the classification of the entire image to one of the classes of interest (plain, longitudinal, transverse, block, or alligator). A large number of features are examined using discriminant analysis, k-Nearest Neighbor, and discrete choice models. Conclusions are drawn on the discriminatory power of the various features and the appropriateness of the classification methods. For both levels, the logit model provided the best and most robust classification results. The sensitivity of the overall classification accuracy to the accuracy of primitive classification is also investigated.

An activity network approach to introducing alternative transportation systems

Abstract: This paper investigates an incremental, cost-effective transition process from an existing transportation system to some improved system. The goal of the process is to speed introduction and transition, compared to gradual evolutionary change, by identifying technologies and feasible pathways. A graphical technique is outlined that combines engineering benefit/cost analysis and precedence diagrams from project management. The approach defines and evaluates potential technologies and relationships among the technologies. The introduction of each technology can be broken into development, implementation, and market growth activities, which are represented on activity networks along with activities for other related technologies. The resulting activity networks can be used to plan and evaluate transition pathways. A market range can be calculated for each technology, based on expected benefits and costs, showing when the technology should be deployed and when it should be replaced. Since each activity has associated time, cost, and resource estimates for completion, scaled activity networks can also be drawn. The most useful diagrams are two-dimensional, with time on the horizontal axis and market level on the vertical axis. These scaled precedence activity network (SPAN) diagrams show the market range for each technology and how soon each technology can be deployed. Logical strategies for introduction, which span the expected market range by using a combination of several technologies, can then be identified from the SPAN diagrams. These deployment strategies, selected by matching the level of deployment of technologies with the market level, provide a means of reducing the investment risk of introducing alternative systems.