Computer supported estimation of input data for transportation models

Abstract

Control and management of transportation systems frequently rely on optimization or simulation methods based on a suitable model. Such a model uses optimization or simulation procedures and correct input data. The input data define transportation infrastructure and transportation flows. Data acquisition is a costly process and so an efficient approach is highly desirable. The infrastructure can be recognized from drawn maps using segmentation, thinning and vectorization. The accurate definition of network topology and nodes position is the crucial part of the process. Transportation flows can be analyzed as vehicle’s behavior based on video sequences of typical traffic situations. Resulting information consists of vehicle position, actual speed and acceleration along the road section. Data for individual vehicles are statistically processed and standard vehicle characteristics can be recommended for vehicle generator in simulation models.