A Passenger-centric Multi-agent System Model for Multimodal Public Transportation

If we want to understand how perturbations spread across a multi-modal public transportation system, we have to include passenger flows into the model and the analysis. Indeed, in general no two different lines in such a system are physically connected directly, or share tracks or other resources. Rather, they are connected by passengers changing lines and thus transmit perturbations from one line or mode to another. We present a formal passenger-centric multi-agent system model that can capture (i) individual and possibly multi-modal trip profiles with branches resulting from different decision outcomes, (ii) the movement of fixed-route operated transportation means, and (iii) in-vehicle and instation capacity constraints. The model is based on a nets-within-nets approach with Petri nets as the basic building entities. Thus, it has a convenient graphical representation, and the possibility of execution.