This paper examines the nature of uncertainty present in transport planning and explores appropriate mathematical treatment. Two types of uncertainty are dealt with: vagueness and ambiguity. The former refers to the uncertainty caused by the lack of definition of words, and the latter refers to the uncertainty caused by the lack of information about the subject matter. The mathematical framework that can deal with vagueness is fuzzy set theory, and that for ambiguity is evidence theory. Differences in the nature of problems that these two types of uncertainty present are examined. How to apply the appropriate mathematical technique to model these types of uncertainty is discussed, along with the basic properties of fuzzy set theory and evidence theory. Domains of transport problems that are conducive to these theories are explained, and the issues that need to be examined when using these theories are discussed.

There is a need for advanced traffic management systems (ATMS) on a local and network level to effectively utilize today's transportation system. An automated operating system, with round the clock incident detection, would help to initiate the required action on time in case of an accident and would minimize the effect of the incident. One of the key elements of ATMS is the vehicle detection unit. The most commonly used inductive loop detectors have a relatively short life span and frequently malfunction. Any maintenance of loop detectors requires road closures. Also, safety of the crew is an issue. An overview of currently available technologies for traffic control and monitoring is presented, with special emphasis on systems that can be installed and maintained without significantly interfering with the traffic. The paper presents the available technologies, provides comparative information about them, and discusses the technical and conceptual problems typically arising during the deployment. The paper describes video detection systems that currently seem to be the most attractive alternative to the traditional intrusive methods. The new perspectives for traffic monitoring and management not available so far are drawn. The typical architecture of these systems is described, and a few popular video systems are presented in some detail. The most critical deployment issues, such as the choice of the equipment, are described, the location of the sensors is discussed, and data transmission problems are considered.

In the past 10 years increased importance has been placed on public participation and environmental concerns in transportation system decision-making. While there are numerous societal objectives to consider when planning and operating a transportation system, it is not clear whether the optimal strategy with respect to one objective is also the optimal strategy with respect to the other objectives. This paper examines how new objectives and environmental considerations can be modeled within the traditional, macroscopic traffic assignment methodology. In addition, a new methodology for modeling the assignment of vehicles in realistic networks is developed based on equitable, rather than user-equilibrium or system-optimal, principles. The basic premise is that with the advent of intelligent transportation systems (ITS) the operation of the transportation system based on the objectives of the general public, rather than the traveling public and system operators, is feasible. A methodology for modeling these situations is required. All of the approaches discussed here were tested on a calibrated network from Edmonton, Alberta, Canada.

The behavior of traffic in the heterogeneous environment of an urban uncontrolled intersection is complex and difficult to model. The present study describes the methodology of simulating the traffic flow and thereby estimating the number of conflicts in varying traffic flow conditions. The arrival pattern of vehicles was represented by a multivariate distribution to generate input to the simulation model. The model was validated externally, using field observed data, and was found to predict the number of conflicts well. As an illustration of usefulness of the model, variation of conflict rate (the probability of a vehicle's getting involved in conflict) due to variation in traffic volume and the proportion of right-turning traffic has been quantified. Under the prevailing traffic composition and turning movements, the conflict rate is estimated to lie in the range of 0.66–0.70, 0.79–0.84, and 0.80–0.87 for intersection volumes of 2,000, 2,500, and 3,000 vehicles per hour, respectively. Issues related to the applicability of the proposed model are briefly discussed.

The objective of this study was to observe how successfully remote-sensing observations can be made on two-lane roads, to identify the factors that affect the success with which observations are made, and to quantify the relationship. The investigation involved both an empirical and a theoretical analysis. The empirical analysis included observation of the number of successful observations made at three sites in Baton Rouge, La., under varying traffic conditions. The theoretical analysis included development of a simulation model of traffic on a two-lane road. It was found that approximately 800 successful observations per hour can be made on two-lane roads in urban areas. This is roughly 80% of the number achievable at single-lane sites. Factors affecting the number of successful observations on two-lane roads are volume, directional split, and traffic composition of the traffic stream. The impact of these traffic characteristics on successful observation was captured in a model that was able to explain 95% of the variation in the data observed in this study.

This paper summarizes a research project that evaluated the performance of traffic marking materials used in Alaska and other northwestern states. This study was conducted by reviewing existing reports, past studies, and information databases; conducting a field survey that rated existing traffic marking; conducting field measurements of retroreflectivity of traffic markings using a reflectometer; and conducting a subjective opinion survey about the performance of traffic markings used in cold regions. Results summarized in this paper include impacts of traffic marking patterns on driver behavior, minimum reflectivity requirements, a general evaluation of traffic marking materials, field survey, reflectivity performance, subjective survey evaluation, and final conclusions. The traffic marking material types evaluated in this study included traffic paint, thermoplastics, preformed tapes, and methyl methacrylate. These traffic marking materials have been applied in Alaska and some northern states.

Headlight sight distance has been an important element in designing sag vertical curves because nighttime drivers need sufficient sight distance for stopping, and also because they depend on the road marking to maintain uniform speed and proper lane positioning. Considering an isolated sag vertical curve in a two-dimensional projection based on the sight distance needs for stopping, the current design standards recommend a minimum length of the sag curve. However, little research work has been directed toward determining the available headlight sight distance in three-dimensional highway alignments. A new analytical model developed by the writers for 3D analysis of headlight sight distance has shown that the 2D analysis may overestimate the available headlight sight distance on sag vertical curves combined with horizontal curves. In this paper, a computer program based on the 3D analytical model is used to evaluate the current 2D standards for sag vertical curves on fill sections. The results show that the current 2D design standards overestimate the available headlight sight distance and therefore underestimate the required length of sag vertical curves combined with horizontal curves. In this paper more accurate values for the required length of sag vertical curves combined with horizontal circular curves are developed. A more economical design is also suggested through the use of asymmetrical (compound) sag vertical curves.

Breakaway supports are common devices used for dual-support signs located along the roadway. The sign system utilizes a fuse plate that is designed to support the wind loads on the sign as well as to allow the sign support to release on impact, and swing up and out of the way, when impacted by an errant vehicle. Two alternative designs for improving the wind load capacity of large dual-support signs were analyzed: a thicker fuse plate design and a balanced hinge design. The safety performance of the sign systems was evaluated using nonlinear, large deformation finite-element analysis (FEA). Crash simulation results indicated that both designs have the potential for meeting the safety recommendations set forth by the National Cooperative Highway Research Program (NCHRP) Report No. 350.

A procedure for identifying critical impact points (CIP) for longitudinal barriers is presented. The procedure is based on computer simulations of longitudinal barrier impacts using the Barrier VII computer program. Critical impact points were selected to maximize wheel snagging and rail tensile loadings. The procedure is based on the finding that the critical impact point is largely dependent on the strength of the posts and the rail bending strength. Thus, the CIP selection procedure relies only on the effective post yield force per unit barrier length and the rail plastic moment. The technique can be applied to all post and beam barriers as well as transitions involving one of these types of barriers. This procedure has been incorporated into the National Cooperative Highway Research Program (NCHRP) Report 350.

This paper introduces a frequency-domain back-calculation program, BKGREEN, that can be used to determine the elastic modulus for each pavement layer. The pavement is modeled as a layered elastic system resting over a rigid bedrock or a half-space in terms of dynamic Green flexibility influence functions. A nonlinear least-squares technique is implemented for matching the measured and calculated deflection response functions. The theoretical results show that (1) the use of multifrequency back-calculation can lead to satisfactory convergence of layer moduli even when the number of receivers is less than the number of layers to be determined; therefore, it overcomes the limitation on the maximum allowable number of layers in most static back-calculation programs; and (2) the program has the potential to cope with difficult situations, such as the existence of extraordinarily thin layers, or shallow hard layers, as well as the degradation of surface layers due to cracking, or temperature and moisture gradients. After studying real field deflections at five test sites, we find that the moduli back-calculated using quasi-static frequency in the BKGREEN program compare well with the other five developed programs.

A formulation for the analysis of a general multilayered pavement system is developed in a manner analogous to that followed for conventional asphalt concrete pavements. The Burmister analytical solution is expanded to the general case of a pavement system in which each member may be treated as a Burmister layer, a Winkler spring bed, or a Kirchhoff plate. The proposed formulation is, therefore, applicable to the interior loading condition of both asphalt and Portland cement concrete pavements, and it has been coded in FORTRAN into the program DIPLOMAT. The program allows the user to idealize pavement members using dissimilar models and may contribute significantly toward the development of a “unified” approach to pavement analysis. The program has been compared with available closed-form solutions, and with computerized layered elastic analysis (the computer program BISAR). Excellent agreement has been found in both instances. Applications of the program DIPLOMAT are also presented.

Temperature is an important factor influencing the functioning of concrete pavements. In the past, analysis of temperature distribution in pavement has been done using numerical techniques such as finite difference or finite element method; both require significant computational efforts. In this paper, a closed-form analytical solution is presented for a three-layer pavement system, subjected to a periodic variation of either ambient air temperature or pavement surface temperature. The thermal analysis is coupled with a plate theory with Winkler foundation to allow for calculation of curling stresses and bending moments. The methods for characterizing periodic temperature variations are also described. The main findings from the numerical studies show that the temperature distribution with depth can be highly nonlinear, particularly when daily temperature variation is considered. Thus, the frequency of temperature variation, rather than the amplitude, has the most significant effect on the calculated temperature distribution with depth in the concrete pavement layer. A frequency of 2π rad/d or higher can be expected to cause nonlinear temperature distribution and require that calculation of curling stresses be based on a nonlinear distribution model.

This technical note presents the graphical version of the solutions of a closed-form back-calculation algorithm for two-layer rigid pavements. The development of the graphs and the theoretical basis for the choice of analysis parameters for the graphical solutions are described. Graphs are developed for back-calculating the moduli of the pavement slab and the subgrade, based on the common deflection measurement data obtainable from typical falling weight deflectometer tests. Two sets of graphs are presented, one for back-calculating the radius of relative stiffness, and the other for determining the deflection factor that relates pavement properties to surface deflections. The graphical solutions are applicable for back-calculating the properties of rigid pavements supported on either a liquid foundation or a solid foundation.