SimSiVIDS: Modelling of an Inductive Sensor for Traffic Applications

Abstract

In this work we develop a model of an inductive loop detector for traffic applications in intelligent transportation systems. Our goal is to work out an appropriate model that allows us to obtain the vehicle inductive signatures by means of a simulator to extract some features and/or study several performances without making use of expensive, not only in time but also in resources, tests in real scenarios using our hardware prototype. As it is shown with the results obtained using both the prototype and the inductive sensor simulator, the vehicle signatures in time and frequency domains exhibit similar characteristics, which validates the model proposed in this work. Moreover, a spectral feature extracted from the signatures in the frequency domain is studied using our software, giving us as a result that such a indicator suffers negligible variations with the vehicle speed or acceleration, but depends on its length i.e., on the vehicle type. This remarkable dependence can be exploited for vehicle classification tasks.