Online Calibration of Microscopic Road Traffic Simulator

Microscopic road traffic simulator is a powerful tool to analyze and evaluate various transportation systems due to its efficiency and risk-free operation. It is, therefore, widely used in traffic engineering field along with the gradual implementation of novel intelligent transportation systems. A reliable microscopic traffic simulator is able to accurately represent the real-world traffic situation when it is effectively calibrated with the combination of field data and proper simulation settings. Based on the existing theoretical calibration framework for the microscopic traffic simulator, this paper proposes an online calibration procedure using genetic algorithm as well as a specific implementation method to provide real-time performance measures that adequately mimic the field traffic situation. The proposed method was tested based on loop detector data demonstrating that real-time traffic modeling can be run in parallel with the real-world traffic process.