Options Matter: Accelerating Multi-Robot Navigation via Changing Safety Filters paper demo

Описание: Control barrier functions (CBFs) are widely used as a minimally invasive technique to augment control commands such that they enforce system safety.
However, this safety guarantee can come at the cost of reduced system performance. Therefore, in this work, we introduce a novel framework for using pre-selected time-invariant CBF candidates to form a time-varying CBF that guarantees safety while maximizing performance of multi-robot navigation tasks. We show that a time-varying convex combination of candidate time-invariant CBFs forms a valid time-varying CBF, which renders the time-dependent safe set forward invariant, and we prove that the safe set is uniformly asymptotically stable. Using this type of time-varying CBFs, an algorithm is introduced that alters the CBF shapes based on a metric of progress to solve a congested navigation task. The performance of the proposed algorithm is validated in hardware experiments. The proposed time-varying CBF selection algorithm achieves up to a 40% reduction in the time required to complete congested navigation tasks compared to a baseline
CBF implementation of each selected time-invariant CBF implementation on its own.