Autonomy-Oriented, Self-Equilibrium Routing Model in Traffic Networks

Abstract

The ultimate goal of routing model in a distributed traffic network is to achieve global traffic flow equilibrium and reduce total travel cost without centralized and complicated control on an ensemble of routing components. Autonomy-oriented computing can provide an autonomy-oriented self-organizing controlling method to coordinate those large-scale computational entities to achieve emergent global behavior based on local interactions and local information. To emerge system equilibrium, a new autonomy-oriented routing model is proposed, which is an integration of dynamic network, routing mechanism, local traffic flow dynamics with a great number of routing agents that asynchronously execute a basic control loop. Based on the basic control loop, traffic flow dynamics, routing agents’ routing choosing and navigated vehicles compose a feedback cycle to let the routing system (routing agents, traffic flow and vehicles) to self-organize. Moreover, a novel self-equilibrium routing schemes, considering distance cost, time delay and traffic capacity simultaneously, is proposed to reduce total time cost for all vehicles in traffic network.