Simulation Model Calibration and Validation: Phase II: Development of Implementation Handbook and Short Course

About the report:

A previous study developed a procedure for microscopic simulation model calibration and validation and evaluated the procedure via two relatively simple case studies using three microscopic simulation models. Results showed that default parameters were unacceptable while calibrated parameters were able to replicate field conditions. Consequently, the study recommended that microscopic simulation models be calibrated and validated before they were used for any evaluations and analyses. A technical review panel determined that the previously developed procedure might not be readily adoptable by Virginia Department of Transportation (VDOT) traffic engineers due to its extensive use of advanced statistical tools and a lack of hands-on case study material. In addition, the proposed procedure needed to be tested with complex network conditions such as urban arterial networks and congested freeway systems. Consequently, it was concluded that what is now needed is (1) a handbook for simulation model calibration and validation that can be easily used by VDOT engineers, and (2) tests of the procedure for various network conditions. The purpose of this project was to develop a handbook for simulation model calibration and validation for VDOT traffic engineers to use for their simulation work and to develop and conduct a hands-on short course to instruct them in the use of the handbook. This study recommended the following: 1.) VDOT traffic engineers should calibrate and validate microscopic simulation models by using the enhanced procedure (i.e., multiple performance measures-based procedure) before using them for any engineering applications to ensure reliable results provided for better decision-making. 2.) When VDOT traffic engineers conduct microscopic simulation model calibration and validation, multiple performance measures collected for multiple days should be used to obtain more reliable results. 3.) Additional hands-on short courses should be offered to expose the procedure to more VDOT traffic engineers and possibly others including consultants who would work for VDOT in the future. VDOT's Learning Center and UVA Center for Transportation Studies Local Technical Assistance Program (LTAP) are possible avenue to offer such short courses. 4.) The handbook should be distributed to VDOT traffic engineers who currently use or plan to use microscopic simulation models for engineering applications. In addition, the prototype program and hands-on short course material (Park, 2006) should be accessible to VDOT traffic engineers.

Authors