Title:Robust Control Barrier Functions for Sampled-Data Systems

Abstract:This paper studies the problem of safe control of sampled-data systems under bounded disturbance and measurement errors with piecewise-constant controllers. To achieve this, we first propose the High-Order Doubly Robust Control Barrier Function (HO-DRCBF) for continuous-time systems where the safety enforcing constraint is of relative degree 1 or higher. We then extend this formulation to sampled-data systems with piecewise-constant controllers by bounding the evolution of the system state over the sampling period given a state estimate at the beginning of the sampling period. We demonstrate the proposed approach on a kinematic obstacle avoidance problem for wheeled robots using a unicycle model. We verify that with the proposed approach, the system does not violate the safety constraints in the presence of bounded disturbance and measurement errors.