Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
32 The potential linking of travel demand forecasting models to traffic microsimulation provides the opportunity to more accurately represent traffic conditions, which can be fed back to choices about travel time, travel route, travel mode, or the decision to travel at all. This section highlights the importance of a feedback mechanism that could incorporate travel time reliability into traditional trip-based travel demand models, emerging activity-based models, and route choice models. The section sum- marizes the implemented synthesis of the research literature and testing of various methods to incorporate travel time reliability into operational travel models. Incor- poration of reliability is primarily considered in the overall framework of demand- network equilibrium, with the demand side represented by an advanced activity-based model (ABM) and the network simulation side represented by an advanced dynamic traffic assignment (DTA). Whenever possible, the discussion is extended to incorporate traditional four-step demand models and static equilibrium assignment models. FINDINGS AND RECOMMENDATIONS ON ABM-DTA INTEGRATION Several aspects of ABM-DTA integration and associated feedback mechanisms are essential and need to be addressed even before incorporation of travel time reliability measures. New methods of equilibration of ABM and DTA include the following tech- nical solutions to be applied in parallel: ⢠Individual schedule consistency and temporal equilibrium. Individual schedule con- sistency means that for each person, the daily schedule (i.e., a sequence of trips and activities) is formed without gaps or overlaps. This solution is based on the fact that a direct integration at the disaggregate level is possible along the temporal dimension if the other dimensions (number of trips, order of trips, and trip destina- tions) are fixed for each individual. The inner loop of temporal equilibrium includes 8 TRAVEL TIME RELIABILITY IN PLANNING MODELS
33 A LEXICON FOR CONVEYING TRAVEL TIME RELIABILITY INFORMATION schedule adjustments in individual daily activity patterns, made because congested travel times are different from planned travel times. The purpose of this feedback is to achieve consistency between generated activity schedules (activity start times, and times and durations) and trip trajectories (trip departure time, duration, and arrival time). The feedback is implemented as part of temporal equilibrium between ABM and DTA when all trip destinations and modes are fixed, but departure times are adjusted until a consistent schedule is built for each individual. In this way, any change in travel time would realistically affect activity durations and vice versa. ⢠Presampling of trip destinations. In the second solution, trip origins, destinations, and departure times are presampled. The DTA process then only needs to pro- duce trajectories for a subset of origins, destinations, and departure times. In this case, the schedule consolidation is implemented through corrections to the de- parture and arrival times (based on the individually simulated travel times) and is employed as an inner loop. The outer loop includes a full regeneration of daily activity patterns and schedules but with a subsample of locations for which many individual trajectories are available. Destinations for which individual trajectories have not been generated at the previous iterations use conventional aggregate originâdestination skims. ⢠Specific methods to ensure equilibration and convergence with individual micro- simulation. These methods include various enforcement and averaging strategies. Enforcement methods are specific to microsimulation and designed to ensure con- vergence of âcrispâ individual choices by suppressing or avoiding Monte Carlo variability. Averaging methods have been borrowed from conventional four-step model ing techniques, but they can be also used with microsimulation as long as they are applied to continuous outputs/inputs such as level of service (LOS) vari- ables and/or synthetic trip tables generated by the demand microsimulation process. ⢠ABM improvements for better compatibility with DTA. Several aspects of ABMs can be improved to provide better inputs to DTA. Such improvements can also avoid additional procedures that are currently applied to overcome structural incompat- ibilities that exist between the two models (e.g., randomly slicing trips by departure time). Three important aspects are (1) enhanced temporal resolution in trip depar- ture choice, (2) car occupancy and associated conversion of person trips into auto trips, and (3) inclusion of route type choice as part of the mode choice tree. ⢠Compatible user segmentation, preserving individual randomized value of time (VOT) and value of reliability (VOR). For full compatibility between the demand model and the network simulation model, the relevant individual parameters have to be transferred between the two. Network simulations, and specifically route choice, are not directly influenced by travel purpose or income or car ownership; these effects can instead be encapsulated in the VOT and VOR parameters. There are two principal ways to ensure the necessary compatibility between ABM and DTA: (1) preserve individual VOT and VOR transferred from ABM to DTA with the cor- responding list of trips to simulate, and (2) form user classes with similar VOT and VOR to simplify path-building procedures that can be applied for each class.
34 A LEXICON FOR CONVEYING TRAVEL TIME RELIABILITY INFORMATION FINDINGS AND RECOMMENDATIONS ON INCORPORATION OF RELIABILITY The incorporation of reliability into a network simulation model requires innovative approaches to generate the reliability measures that are fed into the demand model, to make route choice sensitive to reliability measures, and to ensure that a realistic correlation pattern is taken into account when route-level measures of reliability are constructed from link-level measures. The four main methods for quantifying reliability and its impacts are as follows: ⢠Perceived highway time by congestion levels. This concept is based on statistical evidence that travelers perceive each minute of travel time with a weight related to the level of congestion. Although segmented by congestion levels in this method, perceived highway time is not a direct measure of reliability, because only average travel time is considered. It can serve, however, as a good instrumental proxy for reliability, since the perceived weight of each minute spent in congestion is in part a consequence of associated unreliability. ⢠Time variability distribution measures (or mean-variance approach). This method has received considerable attention in recent years and is considered the most prac- tical direct approach. It assumes that several independent measurements of travel time are known that allow for forming the travel time distribution and the cal- culation of derived measures such as buffer time. One important technical detail with respect to generation of travel time distributions is this: Even if the link-level time variations are known, synthesizing the O-Dâlevel time distribution ( reliability âskimsâ) is a nontrivial task because of the dependence of travel times across adjacent links due to a mutual traffic flow. ⢠Schedule delay cost. According to this concept, the direct impact of travel time unreliability is measured through cost functions (penalties expressed in monetary terms) of being late (or early) compared with the planned schedule of the activity. This approach assumes that the desired schedule (preferred arrival time for each trip) is known for each person and activity in the course of the modeling. This assumption, however, is difficult to meet in a practical model setting. ⢠Loss of activity participation utility. This method can be thought of as a general- ization of the schedule delay concept. It assumes that each activity has a certain temporal utility profile and that individuals plan their schedules to achieve maxi- mum total utility over the modeled period (e.g., day), taking into account expected (average) travel times. Then, any deviation from the expected travel time due to unreliability can be associated with a loss of participation in the corresponding activity (or gain if travel time proved to be shorter). This approach was recently adopted in several research works on DTA formulation integrated with activity scheduling analysis. Similar to the schedule delay concept, however, this approach suffers from data requirements that are difficult to meet in practice. The added complexity of estimation and calibration of all temporal utility profiles for all pos- sible activities and person types is also significant. These concerns make it unreal- istic to adopt this approach as the main concept for the current project.
35 A LEXICON FOR CONVEYING TRAVEL TIME RELIABILITY INFORMATION The main features of the four approaches and associated features that have to be added to the demand model and network simulation model are summarized in Table 8.1. TABLE 8.1. SUMMARY OF METHODS FOR INCORPORATING RELIABILITY INTO TRAVEL MODELS Method Demand Model Network Simulation Perceived highway time Segmentation of highway time by congestion levels with differential weights; no significant modification of model structure required Segmentation of highway time skims by congestion levels; no significant modification of model structure required Mean-variance (time distribution measures) Add variance or standard deviation as LOS variable along with mean travel time and cost to mode choice and other travel choices Add variance or standard deviation to route generalized cost along with mean travel time and cost; employ path-based assignment and/or multiple-run framework; generate route variance or standard deviation skims for demand model Schedule delay cost Specify preferred arrival time (PAT) for each trip exogenously or generate PAT endogenously in time-of-day choice; calculate schedule delay cost based on PAT and travel time distribution Incorporate schedule delay cost into joint route and departure time choice; generate O-D travel time distributions in single-run or multiple-run frameworks Temporal activity profiles for participation in activity Calculate generalized cost, including loss in activity participation based on travel time distribution Incorporate temporal activity profiles into joint route and departure time choice; generate O-D travel time distributions in single-run or multiple-run frameworks FINDINGS AND RECOMMENDATIONS ON IMPLEMENTATION FRAMEWORK The corresponding technical solutions are broken into two groups: single-run frame- work and multiple-run framework. Incorporation of reliability factors into the models can be done in either of two principal ways: ⢠Implicitly in a single model run, in which travel time is implicitly treated as a random variable and its distribution, or some parameters of this distribution (such as mean and variance), are described analytically and used in the modeling process. ⢠Explicitly through multiple runs (scenarios), in which the travel time distribution is not parameterized analytically but is simulated directly or explicitly through multiple model runs with different input variables. The Scenario Manager is an essential tool to make the multiple-run approach operational. There are pros and cons associated with each method. The vision emerging from this research is that both methods are useful, and each could be hybridized to account for different sources of travel time variation in the most adequate and computation- ally efficient way. Whenever possible, single-run analytical methods are considered; they are generally preferable from a theoretical point of view, particularly for network
36 A LEXICON FOR CONVEYING TRAVEL TIME RELIABILITY INFORMATION equilibrium formulations, and in terms of a more efficient use of computational resources in application. Generally, the factors that can be described by means of ana- lytical tools and probabilistic distributions relate to the baseline demand and capacity estimates, day-to-day variability in travel demand, impact of weather conditions, traffic control, route choice, mesoscopic effects associated with traffic flow physics, and individual driver behavior. Factors that can be better modeled through explicit scenarios, rather than captured by probabilistic distributions, mostly relate to special events, road works, and occurrence of incidents. Some factors (like day-to-day fluctua- tions in demand, weather conditions, and traffic control) can be modeled both ways.
