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.     

Hubris or humility? Accuracy issues for the next 50 years of travel demand modeling

This study reviews the 50-year history of travel demand forecasting models, concentrating on their accuracy and relevance for public decision-making. Only a few studies of model accuracy have been performed, but they find that the likely inaccuracy in the 20-year forecast of major road projects is ±30 % at minimum, with some estimates as high as ±40–50 % over even shorter time horizons. There is a significant tendency to over-estimate traffic and underestimate costs, particularly for toll roads. Forecasts of transit costs and ridership are even more uncertain and also significantly optimistic. The greatest knowledge gap in US travel demand modeling is the unknown accuracy of US urban road traffic forecasts. Modeling weaknesses leading to these problems (non-behavioral content, inaccuracy of inputs and key assumptions, policy insensitivity, and excessive complexity) are identified. In addition, the institutional and political environments that encourage optimism bias and low risk assessment in forecasts are also reviewed. Major institutional factors, particularly low local funding matches and competitive grants, confound scenario modeling efforts and dampen the hope that technical modeling improvements alone can improve forecasting accuracy. The fundamental problems are not technical but institutional: high non-local funding shares for large projects warp local perceptions of project benefit versus costs, leading to both input errors and political pressure to fund projects. To deal with these issues, the paper outlines two different approaches. The first, termed ‘hubris’, proposes a multi-decade effort to substantially improve model forecasting accuracy over time by monitoring performance and improving data, methods and understanding of travel, but also by deliberately modifying the institutional arrangements that lead to optimism bias. The second, termed ‘humility’, proposes to openly quantify and recognize the inherent uncertainty in travel demand forecasts and deliberately reduce their influence on project decision-making. However to be successful either approach would require monitoring and reporting accuracy, standards for modeling and forecasting, greater model transparency, educational initiatives, coordinated research, strengthened ethics and reduction of non-local funding ratios so that localities have more at stake.     

Local online kernel ridge regression for forecasting of urban travel times

Accurate and reliable forecasting of traffic variables is one of the primary functions of Intelligent Transportation Systems. Reliable systems that are able to forecast traffic conditions accurately, multiple time steps into the future, are required for advanced traveller information systems. However, traffic forecasting is a difficult task because of the nonlinear and nonstationary properties of traffic series. Traditional linear models are incapable of modelling such properties, and typically perform poorly, particularly when conditions differ from the norm. Machine learning approaches such as artificial neural networks, nonparametric regression and kernel methods (KMs) have often been shown to outperform linear models in the literature. A bottleneck of the latter approach is that the information pertaining to all previous traffic states must be contained within the kernel, but the computational complexity of KMs usually scales cubically with the number of data points in the kernel. In this paper, a novel kernel-based machine learning (ML) algorithm is developed, namely the local online kernel ridge regression (LOKRR) model. Exploiting the observation that traffic data exhibits strong cyclic patterns characterised by rush hour traffic, LOKRR makes use of local kernels with varying parameters that are defined around each time point. This approach has 3 advantages over the standard single kernel approach: (1) It allows parameters to vary by time of day, capturing the time varying distribution of traffic data; (2) It allows smaller kernels to be defined that contain only the relevant traffic patterns, and; (3) It is online, allowing new traffic data to be incorporated as it arrives. The model is applied to the forecasting of travel times on London’s road network, and is found to outperform three benchmark models in forecasting up to 1 h ahead.     

Hybrid artificial neural network and locally weighted regression models for lane-based short-term urban traffic flow forecasting

ABSTRACT

In recent years, there has been considerable research interest in short-term traffic flow forecasting. However, forecasting models offering a high accuracy at a fine temporal resolution (e.g. 1 or 5?min) and lane level are still rare. In this study, a combination of genetic algorithm, neural network and locally weighted regression is used to achieve optimal prediction under various input and traffic settings. The genetically optimized artificial neural network (GA-ANN) and locally weighted regression (GA-LWR) models are developed and tested, with the former forecasting traffic flow every 5-min within a 30-min period and the latter for forecasting traffic flow of a particular 5-min period of each for four lanes of an urban arterial road in Beijing, China. In particular, for morning peak and off-peak traffic flow prediction, the GA-ANN 5-min traffic flow model results in average errors of 3–5% and most 95th percentile errors of 7–14% for each of the four lanes; for the peak and off-peak time traffic flow predictions, the GA-LWR 5-min traffic flow model results in average errors of 2–4% and most 95th percentile errors are lower than 10% for each of the four lanes. When compared to previous models that usually offer average errors greater than 6–15%, such empirical findings should be of interest to and instrumental for transportation authorities to incorporate in their city- or state-wide Advanced Traveller Information Systems (ATIS).     

Forecasting telecommuting

Transportation planners increasingly recognize telecommuting as an important trend. But while they often advocate telecommuting as a transportation demand management strategy, transportation planners have made little progress toward incorporating telecommuting into transportation forecasts, at least partly because of the limited data available. In this paper we explore four alternative methodologies for forecasting telecommuting and discuss the kinds of data that must be collected before these methodologies can be applied. The first approach is trend extrapolation, using curves of technological substitution. Sufficient data are currently available to produce forecasts, albeit highly uncertain forecasts, using this approach. However, even with better data this approach does not address underlying factors and trends that will affect the future of telecommuting. As a result, we explore three additional approaches that should produce more reliable forecasts but which require new data and knowledge about telecommuting: analyzing the characteristics of telecommuters in contrast to nontelecommuters, analyzing factors affecting the individual choice to telecommute, and incorporating telecommuting into traditional transportation forecasting models.     

An integrated multi‐objective model to determine the optimal rescue path and traffic controlled arcs for disaster relief operations under uncertainty environments

This study proposes an integrated multi‐objective model to determine the optimal rescue path and traffic controlled arcs for disaster relief operations under uncertainty environments. The model consists of three sub‐models: rescue shortest path model, post‐disaster traffic assignment model, and traffic controlled arcs selection model to minimize four objectives: travel time of rescue path, total detour travel time, number of unconnected trips of non‐victims, and number of police officers required. Since these sub‐models are inter‐related with each other, they are solved simultaneously. This study employs genetic algorithms incorporated with traffic assignment and K‐shortest path methods to determine optimal rescue path and controlled arcs. To cope with uncertain information associated with the damaged network, fuzzy system reliability theory (weakest t‐norm method) is used to measure the access reliability of rescue path. To investigate the validity and applicability of the proposed model, studies on an exemplified case and a field case of Chi‐Chi earthquake in Taiwan are conducted. The performances of three rescue strategies: without traffic control, selective traffic control (i.e. the proposed model) and absolute traffic control are compared. The results show that the proposed model can maintain the efficiency of rescue activity with minimal impact to ordinary trips and number of police officers required.     

A spatiotemporal correlative k-nearest neighbor model for short-term traffic multistep forecasting

The k-nearest neighbor (KNN) model is an effective statistical model applied in short-term traffic forecasting that can provide reliable data to guide travelers. This study proposes an improved KNN model to enhance forecasting accuracy based on spatiotemporal correlation and to achieve multistep forecasting. The physical distances among road segments are replaced with equivalent distances, which are defined by the static and dynamic data collected from real road networks. The traffic state of a road segment is described by a spatiotemporal state matrix instead of only a time series as in the original KNN model. The nearest neighbors are selected according to the Gaussian weighted Euclidean distance, which adjusts the influences of time and space factors on spatiotemporal state matrices. The forecasting accuracies of the improved KNN and of four other models are compared, and experimental results indicate that the improved KNN model is more appropriate for short-term traffic multistep forecasting than the other models are. This study also discusses the application of the improved KNN model in a time-varying traffic state.     

Integrated model for traffic flow forecasting under rainy conditions

Inclement weather, such as heavy rain, significantly affects road traffic flow operation, which may cause severe congestion in road networks in cities. This study investigates the effect of inclement weather, such as rain events, on traffic flow and proposes an integrated model for traffic flow parameter forecasting during such events. First, an analysis of historical observation data indicates that the forecasting error of traffic flow volume has a significant linear correlation with mean precipitation, and thus, forecasting accuracy can be considerably improved by applying this linear correlation to correct forecasting values. An integrated online precipitation‐correction model was proposed for traffic flow volume forecasting based on these findings. We preprocessed precipitation data transformation and used outlier detection techniques to improve the efficiency of the model. Finally, an integrated forecasting model was designed through data fusion methods based on the four basic forecasting models and the proposed online precipitation‐correction model. Results of the model validation with the field data set show that the designed model is better than the other models in terms of overall accuracy throughout the day and under precipitation. However, the designed model is not always ideal under heavy rain conditions. Copyright © 2016 John Wiley & Sons, Ltd.     

Forecasting containerized traffic for the port of Montreal (1981–1995)

This paper presents an approach to making long-term forecasts of overseas container traffic in the port of Montreal. The paper discusses, first, why performance models explaining traffic variations in ports when postulating a fixed hinterland may not be appropriate for the port of Montreal. Then, the forecasting approach itself is presented. Using 1981 as the base year, projections of container traffic through the port of Montreal until 1995 are developed by considering separately the traffic flows associated with the Canadian and the United States hinterlands. Our approach takes account of anticipated changes in Canadian and U.S. international trade volume, containerization rates and regional growth within Canada. It yields detailed forecasts of containerized traffic, by origin and destination, for 78 commodity groups, 7 world regions and 11 North American regions (i.e. the 10 Canadian provinces plus the United States). Two forecasting scenarios are considered and the aggregate results for 1995 are reported. Finally, a comparison of forecasts with recent data available for 1984 is made, suggesting that the internation trade projections used to generate our forecasts may be too conservative.     

Visco‐elastic traffic flow model

To increase our understanding of the operations of traffic system, a visco‐elastic traffic model was proposed in analogy of non‐Newtonian fluid mechanics. The traffic model is based on mass and momentum conservations, and includes a constitutive relation similar to that of linear visco‐elastic fluids. The further inclusion of the elastic effect allows us to describe a high‐order traffic model more comprehensively because the use of relaxation time indicates that vehicle drivers adjust their time headway in a reasonable and safe range. The self‐organizing behaviour is described by introducing the effects of pressure and visco‐elasticity from the point of view in fluid mechanics. Both the viscosity and elasticity can be determined by using the relaxation time and the traffic sound speed. The sound speed can be approximately represented by the road operational parameters including the free‐flow speed, the jam density, and the density of saturation if the jam pressure in traffic flows is identical to the total pressure at the flow saturation point. A linear stability analysis showed that the traffic flow should be absolutely unstable for disturbances with short spatial wavelengths. There are two critical points of regime transition in traffic flows. The first point happens at the density of saturation, and the second point occurs at a density relating on the sound speed and the fundamental diagram of traffic flows. By using a triangular form flow–density relation, a numerical test based on the new model is carried out for congested traffic flows on a loop road without ramp effect. The numerical results are discussed and compared with the result of theoretical analysis and observation data of traffic flows. Copyright © 2011 John Wiley & Sons, Ltd.     

International experience with speed limits during and prior to the energy crisis of 1973–4

This paper presents an artificial neural network (ANN) based method for estimating route travel times between individual locations in an urban traffic network. Fast and accurate estimation of route travel times is required by the vehicle routing and scheduling process involved in many fleet vehicle operation systems such as dial‐a‐ride paratransit, school bus, and private delivery services. The methodology developed in this paper assumes that route travel times are time‐dependent and stochastic and their means and standard deviations need to be estimated. Three feed‐forward neural networks are developed to model the travel time behaviour during different time periods of the day‐the AM peak, the PM peak, and the off‐peak. These models are subsequently trained and tested using data simulated on the road network for the City of Edmonton, Alberta. A comparison of the ANN model with a traditional distance‐based model and a shortest path algorithm is then presented. The practical implication of the ANN method is subsequently demonstrated within a dial‐a‐ride paratransit vehicle routing and scheduling problem. The computational results show that the ANN‐based route travel time estimation model is appropriate, with respect to accuracy and speed, for use in real applications.     

Equilibrium analysis of the scheduling of tours in congested networks

In this paper, we develop a model of travel in tours that joins several locations by travel through a congested network. We develop a microscopic analysis in continuous time of individual benefits obtained by spending time at each of the locations and costs incurred through travel between them. This is combined with a continuous time macroscopic equilibrium model of travel during congested peak periods to show how individuals' travel choices are influenced by the congestion that result from corresponding choices made by others. We show how different travellers can achieve identical net utilities by making different combinations of choices within the equilibrium. The resulting model can be used to investigate the effect on travel behaviour and individual utility of various transport interventions, and we illustrate this by considering the effect of a peak‐period charge that eliminates congestion.     

Multi‐level traffic control at large four‐leg roundabouts

Yield control and full signalization are typical traffic control solutions that can be used at large roundabouts. In the face of increasing congestion issues, it is preferred to use yield control during off‐peak periods and full signalization during peak periods. To automatically accommodate time‐varying vehicular demands, a multi‐level traffic control (MTC) is developed to implement hybrid yield control and fully actuated control at large four‐leg roundabouts. With new application of traffic control devices and traffic detection system, the right‐of‐way can be assigned to entering and circulating vehicles in three modes. The ‘all entering’ mode is equivalent to a yield control. The ‘no entering’ and ‘concurrent entering’ modes are equivalent to a fully actuated control. On the basis of time headways and occupancy times that are detected on the entry and circulatory roadways, the mode of right‐of‐way assignment can be changed in response to actual traffic conditions. For a specific mode of right‐of‐way assignment, traffic signal operation is managed by some detectable traffic events that are happening. The results of the simulation experiments conducted by VISSIM indicated that: (i) MTC was stabilized at the ‘all entering’ mode during off‐peak periods and at the ‘concurrent entering’ mode during peak periods; (ii) MTC would typically change the mode of right‐of‐way assignment according to actual traffic conditions as vehicular demands increased from off‐peak to peak or decreased from peak to off‐peak; and (iii) statistically speaking, MTC inherited the operational advantages of yield control and fully actuated control, and could be effective in improving the operational performance of large four‐leg roundabouts for all hours of the day, regardless of the level of left‐turn ratios. Copyright © 2016 John Wiley & Sons, Ltd.     

Integrated expression method for technical mapping of traffic parameters using RGB color model

In the field of traffic flow, speed, density, time, and distance are fundamental variables analyzed to predict traffic conditions. Reliable sources of information are gauged using tested mathematical approaches that have been developed. However, a fundamental diagram that could serve as a basis for expression techniques has not been devised. Red–green–blue (RGB) color modeling was used to overcome this limitation in traffic flow. The purpose of this study is to provide a way to understand traffic flow conditions based on features of three traffic flow elements simultaneously. The limitation of three‐dimensional expressions in two‐dimensional paper was extended to multi‐dimensional information. Information on speed, density, and flow were combined into a single RGB color and given the name RGB flow‐density space time‐distance space. This cancels out the effect of each individual's vehicular trajectories and contains five major components of a specific road section. The new gizmo aims to provide information on traffic flow conditions in transition and to stimulate further approaches related to the predictions and understanding of traffic flow. Copyright © 2016 John Wiley & Sons, Ltd.     

Measuring inaccuracy in travel demand forecasting: methodological considerations regarding ramp up and sampling

Project promoters, forecasters, and managers sometimes object to two things in measuring inaccuracy in travel demand forecasting: (1) using the forecast made at the time of making the decision to build as the basis for measuring inaccuracy and (2) using traffic during the first year of operations as the basis for measurement. This paper presents the case against both objections. First, if one is interested in learning whether decisions about building transport infrastructure are based on reliable information, then it is exactly the traffic forecasted at the time of making the decision to build that is of interest. Second, although ideally studies should take into account so-called demand “ramp up” over a period of years, the empirical evidence and practical considerations do not support this ideal requirement, at least not for large-N studies. Finally, the paper argues that large samples of inaccuracy in travel demand forecasts are likely to be conservatively biased, i.e., accuracy in travel demand forecasts estimated from such samples would likely be higher than accuracy in travel demand forecasts in the project population. This bias must be taken into account when interpreting the results from statistical analyses of inaccuracy in travel demand forecasting.     

Using repeated cross-sectional travel surveys to enhance forecasting robustness: Accounting for changing mode preferences

Transportation system capacity and performance, urban form and socio-demographics define the influences and constraints conditioning the preferences of urban residents for different transport modes. Changes in characteristics of urban areas are likely to lead to changes in preferences for alternative modes of transport over time; as a consequence, statistical models to forecast mode choice need to be sensitive to both purposeful changes to urban systems as well as exogenous shocks. We make use of the 1996, 2001 and 2006 household surveys conducted in the Greater Toronto and Hamilton Area to study mode preference evolution and model forecasting performance. These repeated cross-sectional household surveys provide an opportunity to investigate aggregate structural changes in commuting mode preferences over time, in a manner sensitive to changes in the urban area. We focus on commuting mode choices because these trips are prime determinants of peak period congestion and peak spreading. We then address how to combine the three cross-sections econometrically in a robust way that allows for use of a single mode choice model across the entire period. Using independent data from 2012, we are able to compare the individual year and combined models in terms of forecasting performance to demonstrate the combined model’s more robust forecasting performance into the future.     

The changing accuracy of traffic forecasts

Researchers have improved travel demand forecasting methods in recent decades but invested relatively little to understand their accuracy. A major barrier has been the lack of necessary data. We compiled the largest known database of traffic forecast accuracy, composed of forecast traffic, post-opening counts and project attributes for 1291 road projects in the United States and Europe. We compared measured versus forecast traffic and identified the factors associated with accuracy. We found measured traffic is on average 6% lower than forecast volumes, with a mean absolute deviation of 17% from the forecast. Higher volume roads, higher functional classes, shorter time spans, and the use of travel models all improved accuracy. Unemployment rates also affected accuracy—traffic would be 1% greater than forecast on average, rather than 6% lower, if we adjust for higher unemployment during the post-recession years (2008 to 2014). Forecast accuracy was not consistent over time: more recent forecasts were more accurate, and the mean deviation changed direction. Traffic on projects that opened from the 1980s through early 2000s was higher on average than forecast, while traffic on more recent projects was lower on average than forecast. This research provides insight into the degree of confidence that planners and policy makers can expect from traffic forecasts and suggests that we should view forecasts as a range of possible outcomes rather than a single expected outcome.

     

Investigating the irregularity of bus routes: highlighting how underlying assumptions of bus models impact the regularity results

A bus route is inherently unstable: when the system is uncontrolled, buses fail to maintain their time‐headways and tend to bunch. Several mathematical bus motion models were proposed to reproduce the bus behavior and assess management strategies. However, no work has established how the choice of a model impacts the irregularity of modeled bus systems, that is, the non‐respect of scheduled headways. Because of this gap, a large body of existing works assumes that the ability of these models to reproduce instability comes only from stochasticity, although the link between stochastic inputs and the level of irregularity remains unknown. Moreover, some recognized phenomena such as a change of travel conditions during a day or delays at signalized intersections are ignored. To address these shortcomings, this paper provides an overview of existing dynamic bus‐focused models and proposes a simple way to classify them. Commonly used deterministic and stochastic models are compared, which allows quantifying the relative influence of stochasticity of each model component on outputs. Moreover, we show that a change in the system equilibrium in a full deterministic system can lead to irregularity. Finally, this paper proposes a refinement of travel time models to account for non‐dynamic signals. In presence of traffic signals, we show that a bus system can be self‐regulated. Especially, these insights could help to calibrate bus model inputs to better reproduce real data. Copyright © 2014 John Wiley & Sons, Ltd.     

Adaptive traffic parameter prediction: Effect of number of states and transferability of models

Traffic parameters can show shifts due to factors such as weather, accidents, and driving characteristics. This study develops a model for predicting traffic speeds under these abrupt changes within regime switching framework. The proposed approach utilizes Hidden Markov, Expectation Maximization, Recursive Least Squares Filtering, and ARIMA methods for an adaptive forecasting method. The method is compared with naive and mean updating linear and nonlinear time series models. The model is fitted and tested extensively using 1993 I-880 loop data from California and January 2014 INRIX data from Virginia. Analysis for number of states, impact of number of states on forecasting, prediction scope, and transferability of the model to different locations are investigated. A 5-state model is found to be providing best results. Developed model is tested for 1-step to 45-step forecasts. The accuracy of predictions are improved until 15-step over nonadaptive and mean adaptive models. Except 1-step predictions, the model is found to be transferable to different locations. Even if the developed model is not retrained on different datasets, it is able to provide better or close results with nonadaptive and adaptive models that are retrained on the corresponding dataset.