Motion planning for autonomous agents requires compliance with a variety of constraints. However, in some situations, all constraints cannot be simultaneously satisfied, e.g., due to the misbehavior of other traffic participants. In such situations, one strategy for guaranteeing a feasible behavior of the system is deliberately ignoring less important constraints while maintaining the satisfaction of more important ones. We address this problem by presenting a novel velocity planner for an autonomous shuttle that violates temporal logic constraints as little as possible in the sense that important constraints are violated last. In particular, we provide an A * -based velocity planner that minimally violates a hierarchically ordered set of constraints formalized in signal temporal logic. We apply our approach to challenging scenarios from the CommonRoad benchmark suite, showing that our proposed method yields easily interpretable and explainable decisions.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.