VISLAB: Uma ferramenta de apoio à calibração de modelos do simulador de tráfego VISSIM

Abstract

Following the evolution of urban traffic microssimulation model’s methods of calibration, highlighting the work of Lacerda (2016), this paper proposes the creation and aplication of a toolfor helping on the effort of calibrating Vissim’s traffic models. Vissim was chosen because is one of the most used softwares, as much in the academic field as in the professional field, for traffic simulation and for having a API that communicates with various languages, in special Python, that is the barebone behind VisLab’s front-end and back-end. The software VisLab was developed and applied to a study case in which was calibrated a small net of Fortaleza, Ceará, making all the required analysis in one tool, eliminating the necessity of third-party programs like Excel, which grants speed and expediency to the process of calibrating a Vissim’s model. It also was used the VisLab’s calibration package on the study case, converging to a small error of 2,23% on the second generation. VisLab’s source code is avaible to all community o utilize and contribute in <https://github.com/matheusferreira195/VisLAB>.