Traffic Grammar and Algorithmic Complexity in Urban Freeway Flow Patterns

Abstract

This paper uses techniques from formal language theory to describe the linear spatial patterns in urban freeway traffic flows in order to understand and analyze "hidden order" in such high volume systems. A method for measuring randomness based on algorithmic entropy is introduced and developed. These concepts are operationalized using Pincus' approximate entropy formulation in an appropriate illustration. These measures, which may be viewed as counterintuitive, are believed to offer robust and rigorous guidance to enhance the overall understanding of efficiency in urban freeway traffic systems. Utilization of such measures should be facilitated by information generated by real time intelligent transportation systems (ITS) technologies and may prove helpful in real time traffic flow management.

Complex spatial networks in application

“…Global statistics representing the complexity of flows can certainly be generated [50], but more usually we are interested in the flows individually, or as patterns. Understanding individual flows in a complex nonlinear system is difficult, even before we get to networks, which are more realistically fuzzy or probabilistic [51].…”