Routing Model Evaluator

Abstract

We expect that the traffic will be almost optimal when the collective behaviour of autonomous vehicles will determine the traffic. The route selection plays an important role in optimizing the traffic. There are different models of the routing problem. The novel intention-aware online routing game model points out that intention-awareness helps to avoid that the traffic generated by autonomous vehicles be worse than the traffic indicated by classical traffic flow models. The models are important, but their applicability in real life needs further investigations. We are building a test environment, where the decision making methods of the different models can be evaluated in almost real traffic. The almost real traffic runs in a well known simulation platform. The simulation platform also provides tools to calculate a dynamic equilibrium traffic assignment. The calculation needs long time and a lot of computing resources. The routing model evaluator contains an implementation of the routing model which determines the routes for the vehicles. The route selections are injected into the simulation platform, and the simulation platform drives the vehicles. The first results of the investigations with the routing model evaluator show that the route selection of the intention-aware routing model will be able to bring the traffic close to a dynamic equilibrium in real time.