EXECUTIVE SUMMARY

Models defined in simulation tools typically entail mathematical relations and algorithmic procedures that take certain variables as input to produce the output variables of interest. The relations and procedures are typically specified up to a set of parameters whose values are estimated for the context at hand. The process of estimating the model parameters, and otherwise determining the values of parameters that govern various aspects of the model to best represent the given area and context is referred to as calibration.

Current practice typically assumes that parameter values obtained using observations of current conditions will remain applicable under different future conditions. However, the values of parameters may be subjected to changes under future conditions. This diminishes the capability of the simulation tools to produce realistic predictions of the impact of various policies and interventions under future conditions. The objective of this study is to present a calibration framework that helps traffic models to produce meaningful and reliable predictions of the system performance under future conditions and contemplated policy/operational interventions. The scope of this study is to provide information for transportation professionals on how to calibrate analytical/simulation models and tools to data that are reflective of the target future conditions.

Chapter 1 defines the problem, its objective, and its challenges. The general approach that some previous studies have utilized to overcome the challenges and develop models that can capture the behavior of simulation agents under future conditions are discussed. Chapter 2 provides an in-depth analysis of previous studies on the calibration of traffic simulation models. Different studies on calibration such as the MULTITUDE (Methods and tools for supporting the Use caLibration and validaTIon of Traffic simUlation moDEls) project are discussed. The chapter includes information on popular fitness functions used for the purpose of calibration. Furthermore, some of the previous studies are discussed that attempted to address the challenge of producing reliable calibrated models. These studies were grouped in order to shape the main components of the proposed calibration framework including scenario-based calibration, consideration of parameter correlation, and use of vehicle trajectories.

To develop a calibration framework that is mindful of future conditions, an assessment is made regarding limitations of the traffic analysis tools and the data they depend on. The tools require many different layers of input data. Some categories of input data are fundamentally more precise than others. Other data may only have high precision when collected or obtained properly. On top of this, the forecasting method(s) being used adds another potential obstacle to the precision that is necessary for calibration. Moreover, the accuracy, precision, and availability of some data are rapidly changing over time. Chapter 3 contains an assessment of the data required for calibrating transportation simulation models and the limitation associated with the data. The role of trajectories, as a major source of understanding the behavior of drivers (and other actors in the transportation network), in the proposed framework is elaborated. In order to account for parameter correlation, a data structure is introduced to classify the model parameters into three different categories based on the level of uncertainty associated with the parameters. Besides the parameter categorization, each parameter should go through a preprocessing and postprocessing phase.

In chapter 4, based on the review of previous studies and the assessment performed on the required data, a gap analysis is conducted to identify a set of Traffic Analysis Tool calibration needs for transportation improvement evaluations and describe the impacts of these gaps. The set of calibration needs is intended to provide a “snapshot” of potential updates that may be required for the current calibration methods. The libraries of parameters needed to support the development of calibrated models are introduced in this chapter.

Chapter 5 articulates the proposed Traffic Analysis Tool calibration methodology/framework; which will be sensitive to, and reflective of, future conditions, no matter how different it is   from the base condition. The proposed methodology/framework will be designed to achieve better analysis validity for a wider range of improvement alternatives. Using the components introduced in previous chapters, a structure for the library of parameters and the method of selecting values from the libraries to generate simulation agents that capture parameter correlations are discussed. Two analysis methods, namely the scenario-based analysis and the robustness-based analysis, are suggested for the purpose of analyzing the simulation inputs and outputs to achieve a calibrated model.

Three different case studies are presented in chapter 6. The case studies emphasize the role of the main component of the framework. Furthermore, they exhibit different types of analysis (scenario- based and robustness-based) associated with the framework. Each case study was analyzed based on the step-by-step approach provided in chapter 7. The information essentially can help transportation professionals to develop robust models that present a realistic image of the study area under future conditions and various policy interventions.

The information provided in this document is intended to identify methodologies users may consider for calibrating traffic analytical/simulation models and tools to analyze data that are reflective of the target future conditions. This document is not intended to convey any endorsement by FHWA of recommended practices for specific applications, and is not intended to override or augment existing FHWA guidance for the calibration, validation, and reasonableness checking of travel and land use forecasting for project development and NEPA processes, which can be found at https://www.environment.fhwa.dot.gov/nepa/Travel_LandUse/travel_landUse_rpt.aspx.

Additionally, FHWA and the Center for Innovative Finance Support have developed a number of spreadsheet-driven modeling and sketch planning tools that may be used by transportation professionals and policy makers to estimate the effects of road pricing strategies on revenue generation, travel behavior, economic activity, and the environment. Descriptions of these tools and links to user guides and the interactive spreadsheets are available at https://www.fhwa.dot.gov/ipd/tolling_and_pricing/tools.