A Run-Time Assurance Approach for Safe Control of a Quadrotor

Abstract This paper provides a comprehensive and experimentally validated demonstration of control-barrier-function (CBF)-based run time assurance (RTA) for quadrotor systems, spanning continuous- and discrete-time formulations and different relative degree dynamics, namely, 1 Degree-of-Freedom (1-DoF) and 2 Degrees-of-Freedom (2-DoF) configurations, for which safety constraints are enforced on the pitch angle and longitudinal position, respectively. The primary controllers used in this work are quadratic optimal controllers, designed to make the drone follow reference signals in the form of square waves. We examine RTA filters embedding a CBF including some extensions such as High Order CBF (HOCBF) and discrete-time CBF (DCBF). The approaches are validated in simulations and demonstrated on a real-world quadrotor platform at the University of Houston, TX. In the experiments, the reference trajectory was purposefully designed to exceed the boundaries of the prescribed safe sets at times, leading the primary controller to command potentially unsafe control actions. The RTA filters, then, adjusted the primary controller input when necessary to ensure that safety constraints were met at all times. This work advances the state of the art by showing how CBF-based RTA schemes can be reliably integrated with standard optimal controllers and deployed on real hardware, highlighting practical tradeoffs between various implementations. The work fills an important gap between CBF and RTA theory and deployable real-world control systems.