Decentralized probabilistic multi-robot collision avoidance using buffered uncertainty-aware Voronoi cells

In this paper, we present a decentralized and communication-free collision avoidance approach for multi-robot systems that accounts both robot localization sensing uncertainties. The relies on the computation of an uncertainty-aware safe region each to navigate among other robots static obstacles in environment, under assumption Gaussian-distributed uncertainty. particular, at time step, construct chance-constrained buffered Voronoi cell (B-UAVC) given specified probability threshold. Probabilistic is achieved by constraining motion be within its corresponding B-UAVC, i.e. between remains below proposed decentralized, communication-free, scalable with number robust robots’ We applied single-integrator, double-integrator, differential-drive robots, general nonlinear dynamics. Extensive simulations experiments team ground vehicles, quadrotors, heterogeneous teams are performed analyze validate approach.