Calibration of traffic micro-simulation models using the simplex algorithm

Abstract

In recent years micro-simulation has become increasingly important in transportation system modeling. An on-going question is whether they adequately represent reality and whether adequate data exists with which to calibrate these models. Fortunately, over the last ten years there has been a rapid deployment of ITS technologies in most urban areas of North America. While the ITS are developed primarily for real-time traffic operations the data typically are archived and is available for traffic micro-simulation calibration. This paper presents a calibration approach for micro-simulation model which uses ITS data and is based on the sequential simplex algorithm. The test bed is a 23 kilometer section of Interstate 10 in Houston, Texas. Two micro-simulation models (CORSIM and TRANSIMS) were calibrated for the AM Peak for two different OD matrices. It was found that the simplex algorithm had better results then either the default values or a simple, manual calibration.