Software for Calibration of Transmodeler traffic microsimulation models

Calibration is an important process in traffic microsimulation modelling since it enables to represent reality more accurately. In this research we present a general methodology for the calibration process and develop a calibration software for Transmodeler traffic microsimulation. This software follows this methodology and has specific methods for its main steps. We propose an experiment design model to select parameters. In addition we used two genetic algorithms to search the optimal values for the parameters. Besides, this software presents an innovative module to explore and visualize the output of the calibration process. In this paper we present the general design of the software which we tested, alongside our methodology, in a case of study in the city of Pereira, Colombia. The results show that it can find optimal values for the parameters and give the user a global view of the process. It indicates that this software has the capability to calibrate traffic microsimulation models.