Methodology for Validating Dynamic Origin–Destination Matrix Estimation Models with Implications for Advanced Traveler Information Systems

This paper describes a methodology for validating online dynamic O–D matrix estimation models using loop detector data in large-scale transportation networks. The simulation procedure focuses on travel aspects related to the collective trip structure of users, including the amount and duration of trips between O–D pairs, trip departure rates, average travel time from each origin and combinations of them. The analysis identifies emerging systematic patterns between these factors and issues related to the model performance, including network scale effects. This procedure aims to enhance the usage of prior O–D information based on, e.g. travel surveys, that are typically used in the estimation process. Moreover, it seeks to integrate the validation of dynamic O–D matrix estimation models with strategies for identifying target population groups for online planning and assessment of real-time travel information services within the context of Advanced Traveler Information Systems (ATIS).     

Estimation of origin‐destination matrices for a multimodal public transit network

This paper presents a procedure for the estimation of origin‐destination (O‐D) matrices for a multimodal public transit network. The system consists of a number of favored public transit modes that are obtained from a modal split process in a traditional four‐step transportation model. The demand of each favored mode is assigned to the multimodal network, which is comprised of a set of connected links of different public transit modes. An entropy maximization procedure is proposed to simultaneously estimate the O‐D demand matrices of all favored modes, which are consistent with target data sets such as the boarding counts and line segment flows that are observed directly in the network. A case study of the Hong Kong multimodal transit network is used to demonstrate the effectiveness of the proposed methodology.     

A traffic flow simulator for short‐term travel time forecasting

This paper presents an off‐line forecasting system for short‐term travel time forecasting. These forecasts are based on the historical traffic count data provided by detectors installed on Annual Traffic Census (ATC) stations in Hong Kong. A traffic flow simulator (TFS) is developed for short‐term travel time forecasting (in terms of offline forecasting), in which the variation of perceived travel time error and the fluctuations of origin‐destination (O‐D) demand are considered explicitly. On the basis of prior O‐D demand and partial updated detector data, the TFS can estimate the link travel times and flows for the whole network together with their variances and covariances. The short‐term travel time forecasting by O‐D pair can also be assessed and the O‐D matrix can be updated simultaneously. The application of the proposed off‐line forecasting system is illustrated by a numerical example in Hong Kong.     

Estimating time dependent O‐D trip tables during peak periods

Intelligent transportation systems (ITS) have been used to alleviate congestion problems arising due to demand during peak periods. The success of ITS strategies relies heavily on two factors: 1) the ability to accurately estimate the temporal and spatial distribution of travel demand on the transportation network during peak periods, and, 2) providing real‐time route guidance to users. This paper addresses the first factor. A model to estimate time dependent origin‐destination (O‐D) trip tables in urban areas during peak periods is proposed. The daily peak travel period is divided into several time slices to facilitate simulation and modeling. In urban areas, a majority of the trips during peak periods are work trips. For illustration purposes, only peak period work trips are considered in this paper. The proposed methodology is based on the arrival pattern of trips at a traffic analysis zone (TAZ) and the distribution of their travel times. The travel time matrix for the peak period, the O‐D trip table for the peak period, and the number of trips expected to arrive at each TAZ at different work start times are inputs to the model. The model outputs are O‐D trip tables for each time slice in the peak period. 1995 data for the Las Vegas metropolitan area are considered for testing and validating the model, and its application. The model is reasonably robust, but some lack of precision was observed. This is due to two possible reasons: 1) rounding‐off, and, 2) low ratio of total number of trips to total number of O‐D pair combinations. Hence, an attempt is made to study the effect of increasing this ratio on error estimates. The ratio is increased by multiplying each O‐D pair trip element with a scaling factor. Better estimates were obtained. Computational issues involved with the simulation and modeling process are discussed.     

A new fuzzy approach to estimate the O–D matrix from link volumes

Abstract

Estimation of the origin–destination (O–D) trip demand matrix plays a key role in travel analysis and transportation planning and operations. Many researchers have developed different O–D matrix estimation methods using traffic counts, which allow simple data collection as opposed to the costly traditional direct estimation methods based on home and roadside interviews.

In this paper, we present a new fuzzy model to estimate the O–D matrix from traffic counts. Since link data only represent a snapshot situation, resulting in inconsistency of data and poor quality of the estimated O–Ds, the proposed method considers the link data as a fuzzy number that varies within a certain bandwidth. Shafahi and Ramezani's fuzzy assignment method is improved upon and used to assign the estimated O–D matrix, which causes the assigned volumes to be fuzzy numbers similar to what is proposed for observed link counts. The shortest path algorithm of the proposed method is similar to the Floyd–Warshall algorithm, and we call it the Fuzzy Floyd–Warshall Algorithm. A new fuzzy comparing index is proposed by improving the fuzzy comparison method developed by Dubois and Prade to estimate and compare the distance between the assigned and observed link volumes. The O–D estimation model is formulated as a convex minimization problem based on the proposed fuzzy index to minimize the fuzzy distance between the observed and assigned link volumes. A gradient-based method is used to solve the problem. To ensure the original O–D matrix does not change more than necessary during the iterations, a fuzzy rule-based approach is proposed to control the matrix changes.     

A dynamic linear model for the estimation of time‐varying origin–destination matrices from link counts

We propose a dynamic linear model (DLM) for the estimation of day‐to‐day time‐varying origin–destination (OD) matrices from link counts. Mean OD flows are assumed to vary over time as a locally constant model. We take into account variability in OD flows, route flows, and link volumes. Given a time series of observed link volumes, sequential Bayesian inference is applied in order to estimate mean OD flows. The conditions under which mean OD flows may be estimated are established, and computational studies on two benchmark transportation networks from the literature are carried out. In both cases, the DLM converged to the unobserved mean OD flows when given sufficient observations of traffic link volumes despite assuming uninformative prior OD matrices. We discuss limitations and extensions of the proposed DLM. Copyright © 2017 John Wiley & Sons, Ltd.     

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.     

Limits and perspectives of effective O–D matrix correction using traffic counts

Correction of the O–D matrix from traffic counts is a classical procedure usually adopted in transport engineering by practitioners for improving the overall reliability of transport models. Recently, Papola and Marzano [Papola, A., Marzano, V., 2006. How can we trust in the O–D matrix correction procedure using traffic counts? In: Proceedings of the 2006 ETC Conference, Strasbourg] showed through laboratory experiments that this procedure is generally unable to provide for effective correction of the O–D matrix. From a theoretical standpoint, this result can be justified by the lower number of (stochastic) equations (independent observed link flows) with respect to the unknowns (O–D flows). This paper first confirms that this represents the main reason for the failure of this procedure, showing that satisfactory correction is generally obtained when the number of equations is greater than the number of unknowns. Then, since this circumstance does not occur in practice, where the number of O–D pairs usually far exceeds the number of link counts, we explore alternative assumptions and contexts, allowing for a proper balance between unknowns and equations. This can be achieved by moving to within-day dynamic contexts, where a much larger number of equations are generally available. In order to bound the corresponding increase in the number of unknowns, specific reasonable hypotheses on O–D flow variation across time slices must be introduced. In this respect, we analyze the effectiveness of the O–D matrix correction procedure in the usually adopted linear hypothesis on the dynamic process evolution of O–D flows and under the assumption of constant distribution shares. In the second case it is shown that satisfactory corrections can be performed using a small number of time slices of up to 3 min in length, leading to a time horizon in which the hypothesis of constant distribution shares can be regarded as trustworthy and realistic.     

Some comments on origin–destination matrix estimation

Estimation of origin–destination (O–D) matrices from link count data is considered. This problem is challenging because the number of parameters to be estimated is typically larger than the number of network links. As a result, it is (usually) impossible to identify a unique optimal estimate of the O–D matrix from mean link traffic counts. However, information from the covariance matrix of link count data collected over a sequence of days can relieve this problem of indeterminacy. This fact is illustrated through a simple example. The use of second-order statistical properties of the data in O–D matrix estimation is then explored, and a class of estimators proposed. Practical problems of model mis-specification are discussed and some avenues for future research outlined.     

Equity based land‐use and transportation problem

Due to additional trip production by land use development, the O‐D travel costs between some O‐D pairs may also change intuitively. This leads to positive and negative impacts on network users traveling between different O‐D pairs. Therefore the equity issue about the benefit distribution gained from the land‐use development problem is raised. This paper proposes an Equity based Land‐Use Transportation Problem (ELUTP) which is intended to examine the benefit distribution among the network users and the resulting equity associated with land‐use development problem in terms of the change of equilibrium O‐D travel cost. In the resulting bi‐level programming model, the upper level sub‐problem maximizes traffic production incorporating equity constraints, while the lower level sub‐problem is a combined trip distribution/assignment user equilibrium problem. Genetic algorithm based method is applied to test the models using an example network.     

A computational model for the probit‐based dynamic stochastic user optimal traffic assignment problem

This paper focuses on computational model development for the probit‐based dynamic stochastic user optimal (P‐DSUO) traffic assignment problem. We first examine a general fixed‐point formulation for the P‐DSUO traffic assignment problem, and subsequently propose a computational model that can find an approximated solution of the interest problem. The computational model includes four components: a strategy to determine a set of the prevailing routes between each origin–destination pair, a method to estimate the covariance of perceived travel time for any two prevailing routes, a cell transmission model‐based traffic performance model to calculate the actual route travel time used by the probit‐based dynamic stochastic network loading procedure, and an iterative solution algorithm solving the customized fixed‐point model. The Ishikawa algorithm is proposed to solve the computational model. A comparison study is carried out to investigate the efficiency and accuracy of the proposed algorithm with the method of successive averages. Two numerical examples are used to assess the computational model and the algorithm proposed. Results show that Ishikawa algorithm has better accuracy for smaller network despite requiring longer computational time. Nevertheless, it could not converge for larger network. Copyright © 2010 John Wiley & Sons, Ltd.     

Statistical inference on transit route‐level origin–destination flows using automatic passenger counter data

We consider inferring transit route‐level origin–destination (OD) flows using large amounts of automatic passenger counter (APC) boarding and alighting data based on a statistical formulation. One critical problem is that we need to enumerate the OD flow matrices that are consistent with the APC data for each bus trip to evaluate the model likelihood function. The OD enumeration problem has not been addressed satisfactorily in the literature. Thus, we propose a novel sampler to avoid the need to enumerate OD flow matrices by generating them recursively from the first alighting stop to the last stop of the bus route of interest. A Markov chain Monte Carlo (MCMC) method that incorporates the proposed sampler is developed to simulate the posterior distributions of the OD flows. Numerical investigations on an operational bus route under a realistic OD structure demonstrate the superiority of the proposed MCMC method over an existing MCMC method and a state‐of‐the‐practice method. Copyright © 2015 John Wiley & Sons, Ltd.     

Data dependent input control for origin–destination demand estimation using observability analysis

In this paper, we address the observability issue of static O–D estimation based on link counts. Unlike most classic observability analyses that relied only on network topological relationships, our analysis incorporates the actual values of input parameters, thus including network operational relations as well. We first analyze possible mathematical properties of an O–D estimation problem with different data input. We then propose a modeling approach based on mixed-integer program for selecting model input that ensures observability and estimation quality. Through establishing a stronger connection between observability analysis and the corresponding estimation problem, the proposed method aims to improve estimation quality while reducing reliance on erroneous data.     

A new lost time estimation method for right‐turn traffic in Japan considering signal phasing and sneakers

Precise estimation of the capacity for right‐turn traffic (comparable to left‐turn traffic in the USA) is of great importance to determine signal phasing schemes at signalized intersections in Japan, where the left‐hand driving rule is valid. However, in most signal timing procedures across the world, the lost time of right‐turn traffic is simply determined by the duration of intergreen intervals and thus lacks considerations of various signal phasing and driver behavior. Meanwhile, sneakers per cycle are usually applied to account for the number of drivers completing right turns during the effective red portion of the clearance‐and‐change intervals. As a result, an initial cycle length must be hypothesized in order to assess the total number of sneakers within the analysis period. Consequently, a time‐consuming iterative calculation process often becomes necessary. Therefore, the present study aims to develop a new lost time estimation method for right‐turn traffic to overcome the aforementioned drawbacks. Lost times of right‐turn traffic under three conventional phasing plans are theoretically formulated on the basis of a time–space diagram and shock‐wave theory. The new method is validated using field data, with case studies of its application in the signal timing procedure. Results indicated that the proposed method is capable of offering more accurate estimation than conventional approaches, which leads to shorter cycle length and simplifies signal timing process by eliminating an iterative check to determine the number of sneakers. Copyright © 2012 John Wiley & Sons, Ltd.     

A method to estimate trip O‐D patterns using a neural network approach

This paper proposes a method which identifies the trip origin‐destination (O‐D) matrix when many pairs of values for the right hand side column (B) and the bottom row (A) of the matrix are given. The method considers B and A as the cause (input) and effect (output) of a system, respectively, and that the O‐D matrix represents the relationship between the cause and the effect. The relationship which satisfies all pairs of the cause and the effect data exactly may not be identified, but, should a general pattern of the relationship exist, it should emerge when many data sets of B and A are given. Two steps are involved in the method: the first step examines if a consistent O‐D pattern exists; if a pattern is found to exist, the second step identifies the values of the elements of the O‐D matrix. The first step is based on the shape of the possibility distributions of the values of the matrix elements. The second step uses a simple back‐propagation neural network. The method is useful to problems that require identification of the cause‐effect relationship when many sets of data for the cause and effect are available, for example, the station‐to‐station travel pattern on a rapid transit line when the total entering and exiting passengers are known at each station for many different days. The model can also be applied to other transportation problems which involve input and output relation.     

A generalized sensor location model for the estimation of network origin–destination matrices

This study examined the network sensor location problem by using heterogeneous sensor information to estimate link-based network origin–destination (O–D) demands. The proposed generalized sensor location model enables different sensors’ traffic monitoring capabilities to be used efficiently and the optimal number and deployment locations of both passive- and active-type sensors to be determined simultaneously without path enumeration. The proposed sensor location model was applied to solve the network O–D demand estimation problem. One unique aspect of the proposed model and solution algorithms is that they provide satisfactory network O–D demand estimates without requiring unreasonable assumptions of known prior information on O–D demands, turning proportions, or route choice probabilities. Therefore, the proposed model and solution algorithms can be practically used in numerous offline transportation planning and online traffic operation applications.     

Apron capacity at hub airports—the impact of wave‐system structure

At hub airports, dominant airlines/alliance coordinate their flights in time with the aim of increasing the number (and quality) of connections, thus producing a wave‐system in traffic schedules. This paper addresses the impact of concentrating aircraft into waves on airport apron capacity. Existing models for apron capacity estimation are based on the number of stands, stand occupancy time, and demand structure, differing between representative groups of aircraft served at an airport. Criteria for aircraft grouping are aircraft type and/or airline and/or type of service (domestic, international, etc.). Modified deterministic analytical models proposed in this paper also take into account the wave‐system parameters, as well as runway capacity. They include the impact of these parameters on the number of flights in wave, stand occupancy time, and consequently apron capacity. Numerical examples illustrate the difference between apron capacity for an origin–destination airport and a hub airport, under the same conditions; utilization of the theoretical apron capacity at a hub airport, given the wave‐system structure; and utilization of the apron capacity at a hub airport when point‐to‐point traffic is allowed to use idle stands. Furthermore, the influence of different assignment strategies for aircraft stands in the case of hub airports is also discussed. Copyright © 2016 John Wiley & Sons, Ltd.     

Hybrid simulated annealing and genetic algorithm for optimizing arterial signal timings under oversaturated traffic conditions

The implementation of system‐wide signal optimization models requires efficient solution algorithms that can quickly generate optimal or near‐optimal signal timings. This paper presents a hybrid algorithm based on simulated annealing (SA) and a genetic algorithm (GA) for arterial signal timing optimization. A decoding scheme is proposed that exploits our prior expectations about efficient solutions, namely, that the optimal green time distribution should reflect the proportion of the critical lane volumes of each phase. An SA algorithm, a GA algorithm and a hybrid SA‐GA algorithm are developed here using the proposed decoding scheme. These algorithms can be adapted to a wide range of signal optimization models and are especially suitable for those optimizing phase sequences with oversaturated intersections. To comparatively evaluate the performance of the proposed algorithms, we apply them to a signal optimization model for oversaturated arterial intersections based on an enhanced cell transmission model. The numerical results indicate that the SA‐GA algorithm outperforms both SA and GA in terms of solution quality and convergence rate. Copyright © 2014 John Wiley & Sons, Ltd.     

Evaluating airlines with slack‐based measures and meta‐frontiers

This study develops a new performance evaluation model, called the meta dynamic network slack‐based measure (MDN‐SBM). This model incorporates the concept of meta‐frontiers to facilitate comparisons of performance of decision making units, while at the same time it generalizes the slack‐based measure (SBM), network SBM, and dynamic SBM. This MDN‐SBM model is capable of dealing with two important special features of the transportation industry: unfavorable accidents and non‐storable goods, i.e. available seat‐kilometers that are wasting assets that lose value completely if unsold before departure. Hence, this generalized model contains higher differentiable capability than all its SBM‐related submodels in the literature. To demonstrate, 35 international airlines with two divisions (production and consumption) in three terms (one‐year time periods from 2007 to 2009) have been analyzed using meta‐performance efficiency measures (ME) for all decision making units and have also been compared by geographical area (Asia‐Pacific, N. America/Europe). The numerical example and comparison validate the proposed MDN‐SBM model and suggest the airlines should put more focus on input resources reduction for productivity improvement. Copyright © 2016 John Wiley & Sons, Ltd.