Intention-Aware Control Based on Belief-Space Specifications and Stochastic Expansion
This paper develops a correct-by-design controller for an autonomous vehicle interacting with opponent vehicles with unknown intentions. We use discrete-valued random variables to model unknown intentions. Based on this, we define an intention-aware control problem for an autonomous vehicle and a collection of opponent agents with epistemic uncertainty. To this end, we focus on a control objective specified in the belief space with temporal logic specifications. From this stochastic control problem, we derive a sound deterministic control problem using stochastic expansion and solve it using shrinking-horizon model predictive control. The solved intention-aware controller allows a vehicle to adjust its behaviors according to its opponents' intentions. It ensures provable safety by restricting the probabilistic risk under a desired level. We show with experimental studies that the proposed method ensures strict limitation of risk probabilities, validating its efficacy in autonomous driving cases. This work provides a novel solution for the risk-aware control of interactive vehicles with formal safety guarantees.
PDF Abstract