Drawing the Line: Basin Boundaries in Safe Petri Nets

Abstract

Attractors of network dynamics represent the long-term behaviours of the modelled system. Understanding the basin of an attractor, comprising all those states from which the evolution will eventually lead into that attractor, is therefore crucial for understanding the response and differentiation capabilities of a dynamical system. Building on our previous results[2] allowing to find attractors via Petri net Unfoldings, we exploit further the unfolding technique for a backward exploration of the state space, starting from a known attractor, and show how all strong or weak basins of attractions can be explicitly computed.