Modelling daily traffic flow distributions by finite mixtures of simple von mises distributions

Abstract

Graphical representation of traffic flows and their accurate modelling is an important topic in intelligent transportation systems, urban planning, and smart environments in general. In this contribution, location–specific daily traffic flows are represented by finite mixtures of Von Mises distributions. The distribution parameters are found by an evolutionary algorithm (differential evolution) that is able to fit data with a high level of accuracy. Statistical models augmented with frequency data are represented by circular plots that can be used as a form of visually appealing and easily understandable fingerprints of the daily traffic flow patterns.