Search

This paper presents a case study of real-time traffic state estimation. The adopted general approach to the design of universal traffic state estimators for freeway stretches is based on stochastic macroscopic traffic flow modeling and extended Kalman filtering, which are outlined in the paper. The reported investigations were conducted by use of eight-hour traffic measurement data collected from a freeway stretch of 4.1 km close to Munich, Germany. Some key issues are carefully investigated, including the tracking capability of the designed traffic state estimator, significance of the online model parameter estimation, sensitivity of the estimator to the initial values of the estimated model parameters as well as to the related noise standard deviation values, and the capability of the estimator to handle biased flow measurements. The achieved results are quite satisfactory. [ABSTRACT FROM AUTHOR]

Observations of freeway traffic flow are usually quite scattered about an underlying curve when plotted versus density or occupancy. Although increasing the sampling intervals can reduce the scatter, whenever an experiment encompasses a rush hour with transitions in and out of congestion, some outlying data stubbornly remain beneath the ‘equilibrium’ curve. The existence of these nonequilibrium points is a poorly understood phenomenon that appears to contradict the simple kinematic wave (KW) model of traffic flow. This paper provides a tentative explanation of the phenomenon, based on experimental evidence. The evidence was a FIFO queue that grew and receded over two detector stations, generating typical flow-density scatter plots at both locations. The locations were far from other interacting traffic streams. The data revealed that a transition zone where vehicles decelerated gradually existed immediately behind the queue. The transition zone was quite wide (about 1 km at both locations), moved slowly (approximately with the ‘shock’ velocity of KW theory), and as a result spent many minutes over each detector station. Disequilibrium flow-density points arose only when the transition zone was over the detectors, suggesting that the transition zone explains their occurrence. The disequilibrium points drifted gradually from one branch of the curve to the other, as KW theory would have predicted if ‘shocks’ had a characteristic width equal to the dimension of the transition zone. Nothing was found in the data to contradict this view. This paper also shows that in our case, if one neglects the shocks' physical dimension, the position of every vehicle can be predicted with KW theory to within approximately five vehicle spacings. Thus, it appears that KW theory can predict rather accurately traffic behavior at the back of FTFO queues, i.e., when the lanes are equally attractive to all drivers. We end with a discussion offering some perspective on how the findings of this paper related to the traffic thinking found in the current literature. [ABSTRACT FROM AUTHOR]

This paper proposes different "dynamic" estimators using time-varying traffic counts to obtain (discrete) time-varying OD flows or average OD flows. All the estimators can combine counts with other available information, e.g., out-dated matrices and surveys, on a general network and can be formulated as optimization problems. In the case of time-varying OD flows, two types of estimators are proposed. Simultaneous estimators produce joint estimates of the whole set of OD matrices, one for each time slice, whereas sequential estimators produce a sequence of OD estimates for successive time slices. A particular type of simultaneous and sequential estimator (namely Generalized Least Squares) was tested on the Italian Brescia-Padua motorway for which "true" OD flows over a whole day were available. Results were generally satisfactory, showing that also in the "no a priori information" case significant estimates could be obtained. [ABSTRACT FROM AUTHOR]

Existing methods of estimating section (link) density on freeways from data provided by electronic presence (loop) detectors typically require extensive knowledge or strong assumptions on prevailing flow conditions, such as homogeneity. Consequently, these methods are known to produce poor estimates in inhomogeneous conditions, or when a priori knowledge of traffic conditions is not available. In this paper, a new data processing approach is presented which estimates density well over a wide range of traffic conditions. It does this by detecting spatially inhomogeneous traffic conditions and compensating the density estimation algorithm appropriately. The data processing algorithm is computationally simple, is not flow-level dependent, does not require any a priori knowledge of traffic conditions on the road and is insensitive to the types of uncertainty found in detector data. The algorithm uses both flow and occupancy data from adjacent detector stations to track the density on a link. A scalar Kalman filter formulation is used to provide the desired density estimate. The simplicity of the filter algorithm is achieved by using a scalar Generalized Likelihood Ratio (GLR) event detection algorithm to compensate the filter for spatially inhomogeneous conditions. Performance of the algorithm is demonstrated with a microscopic freeway simulation. [ABSTRACT FROM AUTHOR]

The goal of this work is to investigate driver car-following patterns on freeways, particularly as a function of traffic flow level, using a headway distribution model. A number of authors have developed "two-component" vehicular headway distribution models that assume vehicles on a road can be divided into two groups according to whether or not they are interacting with the vehicle ahead. A model of this type, the "semi-Poisson" model proposed by Buckley, is applied to a data base consisting of 42,000 observed headways from a single lane of an urban freeway over a range of flow from 900 to 2000 vehicles per lane per hour. A previously developed computational method allows the distribution of followers headways to be calculated directly from the observed total headway distribution by numerically solving an integral equation without introducing a parametric form for the followers distribution. The resulting followers headway distribution is found to be independent of the flow with a mean of 1.32 s and a standard deviation of 0.52 s. No statistically significant discrepancies are found between the model results and the observed data. The theoretical basis for the semi-Poisson model is discussed and compared with those of other models in order to assess the plausibility of the interpretation with respect to car following. [ABSTRACT FROM AUTHOR]

A hypothetical transportation corridor consists of two parallel routes. One route has properties typical of a congested freeway, the other is considered to be the composite of all alternative routes on city streets. Transfer between the two routes is permitted only at discrete points (entrance or exit ramps). At these points, various methods of traffic control are possible. It is postulated that the control objective is to minimize the total cost of travel to all trips, subject to the hypothesis that travelers choose the route of minimum personal cost. Emphasis is placed on two control methods; metering the flow at entrance ramps, or closing the ramps. Two specific models are considered. In this study (Part I), back- tracking is forbidden but the origin-destination characteristics are fairly general. In a second study (Part 11), backtracking is permitted but the origin-destination characteristics have translational symmetry. In both cases it is found that, if the queuing costs are included in the total travel cost, ramp closure is often a more efficient method of control than ramp metering. [ABSTRACT FROM AUTHOR]

As an idealization of a long freeway imbedded in a network of parallel streets, we consider a hypothetical freeway with equally spaced entrance and exit ramps joining an adjacent slower speed parallel route. The origin-destination pattern of trips is assumed to be translationally symmetric on the adjacent route. Each ramp can be metered (causing a queue to form), closed, or left uncontrolled. Two extreme strategies are considered. In the first, the ramps have a spacing short compared with the average trip length, but all ramps are metered at the same rate. In the second, the spacing between ramps is adjusted by closing ramps, but those which are open are not metered. Flow on the freeway is reduced because travelers will (on the average) travel a certain fraction of their journey on the alternate route, enroute to a ramp. The latter strategy typically gives a lower total travel cost, for flows on the freeway close to its capacity. [ABSTRACT FROM AUTHOR]

An interactive computer-based model is described which permits a reasonable assessment of the decrease in the number of vehicle miles driven when one or several lanes of a freeway are devoted to the exclusive use of "carpools": vehicles carrying at least a specified number of people. The vehicle miles driven are obtained by finding the equilibrium distribution of pool sizes as a function of distance driven, on both regular and priority lanes. A simple but believable economic model incorporating the cost of time and the "utility" of having a vehicle for one's exclusive use is the basis for determining what individuals will do in their self-interest. The economic model is partly verified using existing pooling data without priority lanes. The idea of the priority lane or lanes is, of course, that they move faster than the regular lanes. However, an attractive feature of the priority lane concept is that, in typical cases, traffic flow is actually improved even for those who choose not to pool. Simple runs of the model show that in some cases it can be better for all concerned if two lanes are reserved for carpools rather than one lane, and if vehicles with two occupants are allowed to use the priority lanes rather than requiring three occupants. A 12.5-mi stretch of the Santa Monica Freeway in the Los Angeles area is used as a computational example for the theory. Further refinements to the model are suggested. [ABSTRACT FROM AUTHOR]

The article presents information on a model implemented to determine traffic signal timing in a province known as Tuah. The model was duly validated for Tuah. The supervising traffic engineers were, however, not impressed and loudly proclaimed their reluctance to engage in further funding until a practical problem could be solved, said practical problem being the congestion at the most heavily congested intersection in Tuah. The problem was that rush-hour traffic could not get off the freeway and through the interchange fast enough to avoid queuing back onto the freeway. Realizing that the two problems were interconnected, the traffic engineers took much data, created the proper input to the computer, and adjusted the signal timing on the simulation model until the traffic showed a near-optimal improvement. One year later, two traffic engineers from the country of Falicornia discovered that the interchange had been modeled incorrectly-that a Markovian assumption in the model of the interchange simply did not hold.