New Calibration and Validation Standards for Travel Demand Modeling

The Florida Department of Transportation (DOT) initiated a study of model calibration and validation standards in 2007. Florida DOT and the Model Task Force (MTF) are responsible for overseeing updates to modeling procedures in Florida. Recent updates have provided a good opportunity to evaluate procedures for calibrating and validating models. This paper describes updated standards for calibrating and validating travel demand forecasting models. This paper is divided into four calibration and validation topics. The first section describes a literature review conducted to identify relevant model calibration and validation guidelines, standards and best practice examples. The discussion on guidelines and standards proposes an optimum level of model validity, including a suggested validation process, recommended model calibration and validation checks, and guidelines and standards for different study applications. The best practices section describes a recommended step-by-step modeling process, followed by a discussion of actions to be taken in response to model validation and calibration errors; guidance on model validation checks by study type; emerging validation issues; transferable parameters; and shifting paradigms in modeling. The paper concludes with a discussion of guidelines for different model applications, along with examples of model application checks, and a list of commonly used metrics for scenario testing. This refined approach to model calibration and validation reflects a wide range of benchmarks and standards, with a goal of enhancing the accuracy of travel demand forecasts.

Language

  • English

Media Info

  • Media Type: DVD
  • Features: Figures; References; Tables;
  • Pagination: 20p
  • Monograph Title: TRB 88th Annual Meeting Compendium of Papers DVD

Subject/Index Terms

Filing Info

  • Accession Number: 01123041
  • Record Type: Publication
  • Report/Paper Numbers: 09-3049
  • Files: TRIS, TRB
  • Created Date: Mar 13 2009 6:36AM