Litcius/Paper detail

Maneuvering with safety guarantees using control barrier functions

Mathias Marley, Roger Skjetne, Erlend A. Basso, Andrew R. Teel

2021IFAC-PapersOnLine19 citationsDOIOpen Access PDF

Abstract

Control barrier functions (CBFs) ensure safety of controlled dynamical systems, by restricting the control inputs to render desired sets forward invariant. In this paper we propose a dynamic guidance scheme for autonomous vehicles, using CBFs to reactively generate an obstacle-free trajectory. By implementing the safety constraints on a kinematic guidance level, rather than on a lower-level control layer, we do not need to account for uncertainty in the ship dynamics explicitly. Moreover, for ships with well-proven and resilient control systems, this is an appropriate interface level, since it does not require modification of lower-level feedback control. The guidance scheme is applied to maneuvering of underactuated ships, using a virtual vessel with unicycle dynamics to trace out a feasible trajectory.

Topics & Concepts

TrajectoryUnderactuationKinematicsComputer scienceScheme (mathematics)Control theory (sociology)ObstacleObstacle avoidanceControl (management)Interface (matter)Dynamical systems theoryControl engineeringEngineeringMobile robotRobotMathematicsArtificial intelligenceParallel computingPhysicsQuantum mechanicsLawAstronomyBubbleMathematical analysisClassical mechanicsPolitical scienceMaximum bubble pressure methodGuidance and Control SystemsAdaptive Control of Nonlinear SystemsRobotic Path Planning Algorithms
Maneuvering with safety guarantees using control barrier functions | Litcius