No abstract

Modern navigation systems warn the user of traffic jams ahead and suggest alternative routes. However, a lemming effect can cause the alternative routes also to become congested, as the system suggests the same route to all users. As such, in an attempt to optimize for the individual driver, the welfare of the traffic network is punished. In this paper we introduce an online and proactive method for collective rerouting recommendations based on real-time data and stochastic optimization. Our system periodically monitors the status of the network to identify potentially congested roads together with vehicles affected by them. The system then uses Uppaal Stratego to perform machine learning and approximate the best rerouting scenarios. As a proof of concept, we build a SUMO model of a representative traffic network. We perform exhaustive experiments considering different traffic loads and different traffic light controllers. Our results are promising, showing considerable improvement in travel times, queue lengths, and CO2 emissions.

No abstract

Colored Petri nets offer a compact and user friendly representation of the traditional P/T nets and colored nets with finite color ranges can be unfolded into the underlying P/T nets, however, at the expense of an exponential explosion in size. We present two novel techniques based on static analysis in order to reduce the size of unfolded colored nets. The first method identifies colors that behave equivalently and groups them into equivalence classes, potentially reducing the number of used colors. The second method overapproximates the sets of colors that can appear in places and excludes colors that can never be present in a given place. Both methods are complementary and the combined approach allows us to significantly reduce the size of multiple colored Petri nets from the Model Checking Contest benchmark. We compare the performance of our unfolder with state-of-the-art techniques implemented in the tools MCC, Spike and ITS-Tools, and while our approach is competitive w.r.t. unfolding time, it also outperforms the existing approaches both in the size of unfolded nets as well as in the number of answered model checking queries from the 2021 Model Checking Contest.