Parameter identification on helly’s car-following model

Abstract

Helly’s Model is a very simple and well-known differential equation that describes the car-following phenomenon very accurately with a very intuitive approach. It involves relative distances and velocities between two cars (one in front of the other) which are the variables of the model and which are related through constants which give proportionality and consistency to it. Those constants are parameters that must be calculated in order to achieve the necessary similarity to the real behaviour that they are modelling. In this work, an identification scheme for such estimation is presented. A set of measured data taken from real driving experiments are used to calculate the values of such parameters. Then, a simulation of the velocity developed by one of the cars is performed in order to compare this simulated behaviour against the data directly measured from that same car. The results show that there the differences of both sets of data are minimal, and that the model is very well adjusted.