View article

While nowadays many articles on the application, calibration and validation of (microscopic) traffic simulation models discuss the need for a certain number of replications to obtain reliable results, none of them seems to discuss clearly how exactly to determine the number required.