Symmetry breaking in optimal transport networks

This repository delves into the optimization of urban transportation networks, focusing on the interplay between road (local) and subway (global) networks. We use simulated annealing and multi-greedy optimization to study optimal subway structures. On analyzing real-world population distributions, distinct patterns emerge: optimal subway networks with one or two central branches. Similar findings were observed for Atlanta and Boston. Furthermore, we probed real subway efficiencies in these cities, comparing them to multi-greedy solutions. Notably, as local speed (a proxy for congestion) increased, real subways deviated more from the optimal.