Reactive coordination rules for traffic optimization in road sharing problems
Abstract
In the context of transportation of goods, autonomous vehicles are considered today as a solution for large platforms. We are interested in managing unexpected events, like failure of a vehicle or presence of obstacles on the road, as they can generate global phenomena and complex traffic congestions (such as traffic jams). We explore solutions to avoid such undesirable emergent behaviors by studying local rules for coordinating agents (vehicles). We focus on managing space sharing conflicts at the local level, i.e. between the involved vehicles. We consider a generic scenario where two queues of vehicles share a single lane. We propose a model of the network as well as the agents, and simple coordination rules that only involve the two vehicles at the front of each queue. We then conduct experiments that allow the analysis and the comparison of the proposed self-regulation rules. We show that the alternating strategy commonly used by drivers can be easily improved to minimize the delay of the different vehicles.
Domains
Artificial Intelligence [cs.AI]
Origin : Publisher files allowed on an open archive
Loading...