A maximum entropy-least squares estimator for elastic origin-destination trip matrix estimation

In transportation subnetwork-supernetwork analysis, it is well known that the origin-destination (O-D) flow table of a subnetwork is not only determined by trip generation and distribution, but also a result from traffic routing and diversion, due to the existence of internal-external, external-internal and external-external flows. This result indicates the variable nature of subnetwork O-D flows. This paper discusses an elastic O-D flow table estimation problem for subnetwork analysis. The underlying assumption is that each cell of the subnetwork O-D flow table contains an elastic demand function rather than a fixed demand rate and the demand function can capture all traffic diversion effect under various network changes. We propose a combined maximum entropy-least squares estimator, by which O-D flows are distributed over the subnetwork in terms of the maximum entropy principle, while demand function parameters are estimated for achieving the least sum of squared estimation errors. While the estimator is powered by the classic convex combination algorithm, computational difficulties emerge within the algorithm implementation until we incorporate partial optimality conditions and a column generation procedure into the algorithmic framework. Numerical results from applying the combined estimator to a couple of subnetwork examples show that an elastic O-D flow table, when used as input for subnetwork flow evaluations, reflects network flow changes significantly better than its fixed counterpart.     

Estimation of origin-destination matrices from link traffic counts on congested networks

Conventional methods for estimating origin-destination (O-D) trip matrices from link traffic counts assume that route choice proportions are given constants. In a network with realistic congestion levels, this assumption does not hold. This paper shows how existing methods such as the generalized least squares technique can be integrated with an equilibrium traffic assignment in the form of a convex bilevel optimization problem. The presence of measurement errors and time variations in the observed link flows are explicitly considered. The feasibility of the model is always guaranteed without a requirement for estimating consistent link flows from counts. A solution algorithm is provided and numerical simulation experiments are implemented in investigating the model's properties. Some related problems concerning O-D matrix estimation are also discussed.     

Norm approximation method for handling traffic count inconsistencies in path flow estimator

Path flow estimator (PFE) is a one-stage network observer proposed to estimate path flows and hence origin–destination (O–D) flows from traffic counts in a transportation network. Although PFE does not require traffic counts to be collected on all network links when inferring unmeasured traffic conditions, it does require all available counts to be reasonably consistent. This requirement is difficult to fulfill in practice due to errors inherited in data collection and processing. The original PFE model handles this issue by relaxing the requirement of perfect replication of traffic counts through the specification of error bounds. This method enhances the flexibility of PFE by allowing the incorporation of local knowledge, regarding the traffic conditions and the nature of traffic data, into the estimation process. However, specifying appropriate error bounds for all observed links in real networks turns out to be a difficult and time-consuming task. In addition, improper specification of the error bounds could lead to a biased estimation of total travel demand in the network. This paper therefore proposes the norm approximation method capable of internally handling inconsistent traffic counts in PFE. Specifically, three norm approximation criteria are adopted to formulate three L_p-PFE models for estimating consistent path flows and O–D flows that simultaneously minimize the deviation between the estimated and observed link volumes. A partial linearization algorithm embedded with an iterative balancing scheme and a column generation procedure is developed to solve the three L_p-PFE models. In addition, the proposed L_p-PFE models are illustrated with numerical examples and the characteristics of solutions obtained by these models are discussed.     

Estimating intersection O-D matrices from observed time-series of traffic counts: The zellner estimator

The problem of estimating intersection O-D matrices from input and output time-series of traffic counts is considered in this paper. Because of possible existence of significant correlation between the error terms across structural equations forming the O-D matrices, the seemingly unrelated estimator (Zellner estimator) was suggested. Estimation results showed evidence of strong correlation between error terms across-equations. Generally, the Zellner estimator produced more efficient estimates than did the ordinary least-squares estimator. Furthermore, the Zellner estimator satisfied all constraints and reproduced turning movements comparable to the actual ones.     

Transport demand model estimation from traffic counts

Many problems in transport planning and management tasks require an origindestination (O-D) matrix to represent the travel pattern. However, O-D matrices obtained through a large scale survey such as home or roadside interviews, tend to be costly, labour intensive and time disruptive to trip makers. Therefore, the use of low cost and easily available data is particularly attractive.The need of low-cost methods to estimate current and future O-D matrices is even more valuable in developing countries because of the rapid changes in population, economic activity and land use. Models of transport demand have been used for many years to synthesize O-D matrices in study areas. A typical example of this is the gravity model; its functional form, plus the appropriate values for the parameters involved, is employed to produce acceptable matrices representing trip making behaviour for many trip purposes and time periods.The work reported in this paper has combined the advantages of acceptable travel demand models with the low cost and availability of traffic counts. Three types of demand models have been used: gravity (GR), opportunity (OP) and gravity-opportunity (GO) models. Three estimation methods have been developed to calibrate these models from traffic counts, namely: non-linear-least-squares (NLLS), weighted-non-linear-least-squares (WNLLS) and maximumlikelihood (ML).The 1978 Ripon (urban vehicle movement) survey was used to test these methods. They were found to perform satisfactorily since each calibrated model reproduced the observed O-D matrix fairly closely. The tests were carried out using two assignment techniques, all-or-nothing and the stochastic method due to Burrell, in determining the routes taken through the network.requests for offprints     

Estimating origin-destination matrices from roadside survey data

The purpose of this study is to develop a valid and efficient method for estimating origin-destination tables from roadside survey data. Roadside surveys, whether conducted by interviews or postcard mailback methods, typically have in common the sampling of trip origin and destination information at survey stations. These survey stations are generally located where roads cross “screenlines,” which are imaginary barriers drawn to intercept the trip types of interest.Such surveys also include counts of traffic volumes, by which the partial origin-destination (O-D) tables obtained at the different stations can be expanded and combined to obtain the complete O-D table which represents travel throughout the entire study area. The procedure used to expand the sample O-D information from the survey stations must recognize and deal appropriately with a number of problems:

• 1.(i) The “double counting” problem: Long-distance trips may pass through more than one survey station location; thus certain trips have the possibility of being sampled and expanded more than once, leading to a potentially serious overrepresentation of long-distance trips in the complete expanded trip table.
• 2.(ii) The “leaky screenline” problem: Some route choices, particularly those using very lightly traveled roads, may miss the survey stations entirely, leading to an underestimation of certain O-D patterns, or to distorted estimates if such sites are arbitrarily coupled with actual nearby station locations.
• 3.(iii) The efficient use of the data: There is a need to adjust expansion factors to compensate for double counting and leaky screenlines. How can this be accomplished such that all of the data obtained at the stations are used without loss of information?
• 4.(iv) The consequences of uncertainty and unknown travel behavior: Since the O-D data and other sampled variables are subject to random error, and since in general the probability of encountering a long-distance trip at some survey stations is affected by traveler route-choice behavior, which is not understood, the sample expansion procedure must rely on the use of erroneous input data and questionable assumptions. The preferred procedure must minimize, rather than amplify, the effects of such input errors.

Here, five alternate methods for expanding roadside survey data in an unbiased manner are proposed and evaluated. In all cases, it is assumed that traveler route choice generally follows the pattern described by Dial's multipath assignment method. All methods are applied to a simple hypothetical network in order to examine their efficiency and error amplification properties. The evaluation of the five methods reveals that their performance properties vary considerably and that no single method is best in all circumstances. A microcomputer program has been provided as a tool to facilitate comparison among methods and to select the most appropriate expansion method for a particular application.     

A new class of dynamic methods for the identification of origin-destination flows

A new systems dynamics approach for the identification of origin-destination (O-D) flows in a traffic system is presented. It is the basic idea of this approach that traffic flow through a facility is treated as a dynamic process in which the sequences of short-time exit flow counts depend by causal relationships upon the time-variable sequences of entrance flow volumes. In that way enough information can be obtained from the counts at the entrances and the exits to obtain unique and bias-free estimates for the unknown O-D flows without further a priori information. Four different methods were developed: an ordinary least squares estimator involving cross-correlation matrices, a constrained optimization method, a simple recursive estimation formula and estimation by Kalman filtering. The methods need only moderate computational effort and are particularly useful for tracking time-variable O-D patterns for on-line identification and control purposes. An analysis of the accuracy of the estimates and a discussion of the convergence properties of the methods are given. Finally, a comparison with some conventional static estimation procedures is carried out using synthetic as well as real data from several intersections. These tests demonstrated that the presented dynamic methods are highly superior to conventional techniques and produce more accurate results.     

Path flow estimator for planning applications in small communities

This paper presents an alternative planning framework to model and forecast network traffic for planning applications in small communities, where limited resources debilitate the development and applications of the conventional four-step travel demand forecasting model. The core idea is to use the Path Flow Estimator (PFE) to estimate current and forecast future traffic demand while taking into account of various field and planning data as modeling constraints. Specifically, two versions of PFE are developed: a base year PFE for estimating the current network traffic conditions using field data and planning data, if available, and a future year PFE for predicting future network traffic conditions using forecast planning data and the estimated base year origin–destination trip table as constraints. In the absence of travel survey data, the proposed method uses similar data (traffic counts and land use data) as a four-step model for model development and calibration. Since the Institute of Transportation Engineers (ITE) trip generation rates and Highway Capacity Manual (HCM) are both utilized in the modeling process, the analysis scope and results are consistent with those of common traffic impact studies and other short-range, localized transportation improvement programs. Solution algorithms are also developed to solve the two PFE models and integrated into a GIS-based software called Visual PFE. For proof of concept, two case studies in northern California are performed to demonstrate how the tool can be used in practice. The first case study is a small community of St. Helena, where the city’s planning department has neither an existing travel demand model nor the budget for developing a full four-step model. The second case study is in the city of Eureka, where there is a four-step model developed for the Humboldt County that can be used for comparison. The results show that the proposed approach is applicable for small communities with limited resources.     

Origin-destination trip table estimation based on subarea network OD flow and vehicle trajectory data

Identifying accurate origin-destination (O-D) travel demand is one of the most important and challenging tasks in the transportation planning field. Recently, a wide range of traffic data has been made available. This paper proposes an O-D estimation model using multiple field data. This study takes advantage of emerging technologies – car navigation systems, highway toll collecting systems and link traffic counts – to determine O-D demand. The proposed method is unique since these multiple data are combined to improve the accuracy of O-D estimation for an entire network. We tested our model on a sample network and found great potential for using multiple data as a means of O-D estimation. The errors of a single input data source do not critically affect the model’s overall accuracy, meaning that combining multiple data provides resilience to these errors. It is suggested that the model is a feasible means for more reliable O-D estimation.     

Designing heterogeneous sensor networks for estimating and predicting path travel time dynamics: An information-theoretic modeling approach

With a particular emphasis on the end-to-end travel time prediction problem, this paper proposes an information-theoretic sensor location model that aims to minimize total travel time uncertainties from a set of point, point-to-point and probe sensors in a traffic network. Based on a Kalman filtering structure, the proposed measurement and uncertainty quantification models explicitly take into account several important sources of errors in the travel time estimation/prediction process, such as the uncertainty associated with prior travel time estimates, measurement errors and sampling errors. By considering only critical paths and limited time intervals, this paper selects a path travel time uncertainty criterion to construct a joint sensor location and travel time estimation/prediction framework with a unified modeling of both recurring and non-recurring traffic conditions. An analytical determinant maximization model and heuristic beam-search algorithm are used to find an effective lower bound and solve the combinatorial sensor selection problem. A number of illustrative examples and one case study are used to demonstrate the effectiveness of the proposed methodology.     

An intersection turning movement estimation procedure based on path flow estimator

Estimation of intersection turning movements is one of the key inputs required for a variety of transportation analysis, including intersection geometric design, signal timing design, traffic impact assessment, and transportation planning. Conventional approaches that use manual techniques for estimation of turning movements are insensitive to congestion. The drawbacks of the manual techniques can be amended by integrating a network traffic model with a computation procedure capable of estimating turning movements from a set of link traffic counts and intersection turning movement counts. This study proposes using the path flow estimator, originally used to estimate path flows (hence origin–destination flows), to derive not only complete link flows, but also turning movements for the whole road network given some counts at selected roads and intersections. Two case studies using actual traffic counts are used to demonstrate the proposed intersection turning movement estimation procedure. Copyright © 2010 John Wiley & Sons, Ltd.     

Computational experience with an application of a simultaneous transportation equilibrium model to urban travel in Austin,Texas

Safwat and Magnanti (1988) have developed a combined trip generation, trip distribution, modal split, and traffic assignment model that can predict demand and performance levels on large-scale transportation networks simultaneously, i.e. a simultaneous transportation equilibrium model (STEM). The major objective of this paper is to assess the computational efficiency of the STEM approach when applied to an urban large-scale network, namely the urban transportation system of Austin, Texas. The Austin network consisted of 520 zones, 19,214 origin-destination (O-D) pairs, 7,096 links and 2,137 nodes. The Central Processing Unit (CPU) time on an IBM 4381 mainframe computer was 430 seconds for a typical iteration and about 4,734 seconds, less than 79 minutes, to arrive at a reasonably accurate solution in no more than 10 iterations. The computational time at any given iteration is comparable to that of the standard fixed demand traffic assignment procedure. These results encourage further applications of the STEM model to large urban areas.     

Modeling measurement errors and missing initial values in freeway dynamic origin–destination estimation systems

Most existing dynamic origin–destination (O–D) estimation approaches are grounded on the assumption that a reliable initial O–D set is available and traffic volume data from detectors are accurate. However, in most traffic systems, both types of critical information are either not available or subjected to some level of measurement errors such as traffic counts and speed measurement from sensors. To contend with those critical issues, this study presents two robust algorithms, one for estimation of an initial O–D set and the other for tackling the input measurement errors with an extended estimation algorithm. The core concept of the initial O–D estimation algorithm is to decompose the target network in a number of sub-networks based on proposed rules, and then execute the estimation of the initial O–D set iteratively with the observable information at the first time interval. To contend with the inevitable detector measurement error, this study proposes an interval-based estimation algorithm that converts each model input data as an interval with its boundaries being set based on some prior knowledge. The performance of both proposed algorithms has been tested with a simulated system, the I-95 freeway corridor between I-495 and I-695, and the results are quite promising.     

Inferences on trip matrices from observations on link volumes: A Bayesian statistical approach

Previous methods for estimating a trip matrix from traffic volume counts have used the principles of maximum entropy and minimum information. These techniques implicitly give as little weight to prior information on the trip matrix as possible. The new method proposed here is based on Bayesian statistical inference and has several advantages over these earlier approaches. It allows complete flexibility in the degree of belief placed on the prior estimate of the trip matrix and also allows for different degrees of belief in diffeent parts of the prior estimate. Furthermore under certain assumptions the method reduces to a simple updating scheme in which observations on the link flows successively modify the trip matrix. At the end of the scheme confidence intervals are available for the estimates of the trip matrix elements.     

The estimation of a time-dependent OD trip table with vehicle trajectory samples

Abstract

In this paper we discuss a dynamic origin–destination (OD) estimation problem that has been used for identifying time-dependent travel demand on a road network. Even though a dynamic OD table is an indispensable data input for executing a dynamic traffic assignment, it is difficult to construct using the conventional OD construction method such as the four-step model. For this reason, a direct estimation method based on field traffic data such as link traffic counts has been used. However, the method does not account for a logical relationship between a travel demand pattern and socioeconomic attributes. In addition, the OD estimation method cannot guarantee the reliability of estimated results since the OD estimation problem has a property named the ‘underdetermined problem.’ In order to overcome such a problem, the method developed in this paper makes use of vehicle trajectory samples with link traffic counts. The new method is applied to numerical examples and shows promising capability for identifying a temporal and spatial travel demand pattern.     

Link choice proportions from trip distribution and assignment models: An overview and comparison

A number of estimation procedures have been suggested for the situation where a prior estimate of an origin-destination matrix is to be updated on the basis of recently-acquired traffic counts. These procedures assume that both the link flows and the proportionate usage of each link made by each origin-destination flow (referred to collectively as the link choice proportions) are known. This paper examines the possibility and methods for estimating the link choice proportions. Three methods are presented: (1) using ad hoc iteration between trip distribution and traffic assignment; (2) combining trip distribution and assignment in one step; (3) solving a new optimization problem in which the path flows are directly considered as variables and its optimal solution is governed by a logit type formula. The algorithms, covergencies and computational efficiencies of these methods are investigated. Results of testing the three methods on example networks are discussed.     

Comparison of results of two models of transportation demand in Hong Kong: CDAM and a version of microTRIPS

CDAM is a new computer program for solving the combined trip distribution and assignment model for multiple user classes, which enables transport planners to estimate consistent Origin-Destination (O-D) matrices and equilibrium traffic flows simultaneously if the trip production and attraction of each user class at zone centroids are available. This paper reports an application of CDAM to the central Kowloon study area in Hong Kong. The coefficients of the model related to the components of generalized costs are calibrated on 1986 travel data. A comparison of results of CDAM and a version of MicroTRIPS models of transportation demand in Hong Kong are presented. Finally, some conclusions are drawn and the advantage of the CDAM are discussed.     

Calibrating aggregate travel demand models with traffic counts: Estimators and statistical performance

Traffic counts on network links constitute an information source on travel demand which is easy to collect, cheap and repeatable. Many models proposed in recent years deal with the use of traffic counts to estimate Origin/Destination (O/D) trip matrices under different assumptions on the type of "a-priori" information available on the demand (surveys, outdated estimates, models, etc.) and the type of network and assignment mapping (see Cascetta & Nguyen 1988). Less attention has been paid to the possibility of using traffic counts to estimate the parameters of demand models. In this case most of the proposed methods are relative to particular demand model structures (e.g. gravity-type) and the statistical analysis of estimator performance is not thoroughly carried out. In this paper a general statistical framework defining Maximum Likelihood, Non Linear Generalized Least Squares (NGLS) and Bayes estimators of aggregated demand model parameters combining counts-based information with other sources (sample or a priori estimates) is proposed first, thus extending and generalizing previous work by the authors (Cascetta & Russo 1992). Subsequently a solution algorithm of the projected-gradient type is proposed for the NGLS estimator given its convenient theoretical and computational properties. The algorithm is based on a combination of analytical/numerical derivates in order to make the estimator applicable to general demand models. Statistical performances of the proposed estimators are evaluated on a small test network through a Monte Carlo method by repeatedly sampling "starting estimates" of the (known) parameters of a generation/distribution/modal split/assignment system of models. Tests were carried out assuming different levels of "quality" of starting estimates and numbers of available counts. Finally NGLS estimator was applied to the calibration of the described model system on the network of a real medium-size Italian town using real counts with very satisfactory results in terms of both parameter values and counted flows reproduction.     

Multi-step prediction of experienced travel times using agent-based modeling

This paper develops an agent-based modeling approach to predict multi-step ahead experienced travel times using real-time and historical spatiotemporal traffic data. At the microscopic level, each agent represents an expert in a decision-making system. Each expert predicts the travel time for each time interval according to experiences from a historical dataset. A set of agent interactions is developed to preserve agents that correspond to traffic patterns similar to the real-time measurements and replace invalid agents or agents associated with negligible weights with new agents. Consequently, the aggregation of each agent’s recommendation (predicted travel time with associated weight) provides a macroscopic level of output, namely the predicted travel time distribution. Probe vehicle data from a 95-mile freeway stretch along I-64 and I-264 are used to test different predictors. The results show that the agent-based modeling approach produces the least prediction error compared to other state-of-the-practice and state-of-the-art methods (instantaneous travel time, historical average and k-nearest neighbor), and maintains less than a 9% prediction error for trip departures up to 60 min into the future for a two-hour trip. Moreover, the confidence boundaries of the predicted travel times demonstrate that the proposed approach also provides high accuracy in predicting travel time confidence intervals. Finally, the proposed approach does not require offline training thus making it easily transferable to other locations and the fast algorithm computation allows the proposed approach to be implemented in real-time applications in Traffic Management Centers.