Improving traffic mathematically: Expanding on the work of Chrisostomi et al. (2011)

Abstract

In their 2011 article, Crisostomi, Kirkland and Shorten introduced a new paradigm for modeling road network dynamics using Markov chains to process data harvested in near real time from the roads. It was shown to work, although only for a limited test case. I used the same method and compared the results with results from traffic simulator SUMO. Comparing the calculated stationary distribution, mean first passage time matrix and Kemeny constant to the results from SUMO will show us how their idea works for larger, more complicated networks as well.