A hybrid kinematic controller for resilient obstacle avoidance of autonomous ships

Abstract

Resilience is an important feature of autonomous systems. To be resilient, a control system must be stable, robust, and safe. This paper explores the use of hybrid feedback controllers to ensure robustness towards uncertainties and disturbances in motion control systems for autonomous ships. Motivated by recent developments in control barrier functions (CBFs) for safe maneuvering of autonomous ships, a CBF-based hybrid kinematic controller for obstacle avoidance is proposed. The controller uses course angle as control input, making it suitable for ships with a limited speed envelope. The performance of the controller is illustrated by simulations, using an underactuated ship as a case study.

A kinematic hybrid feedback controller on the unit circle suitable for orientation control of ships

“…However, changing the direction of motion in such a manner is unpredictable for neighboring ships. For autonomous ships, decisiveness and clearly showing the intentions is a desired property [22]. Decisiveness is an inherent property of the hybrid control system presented herein.…”

Sufficient Conditions for Uniform Asymptotic Stability and Input-to-State Stability Using High-Order Control Barrier Functions

“…The convexity and outer semicontinuity properties stated in Lemma 2 are used in the proof of Proposition 1 below, which establishes nominal robustness for discontinuous safeguarding control laws designed using CBFs. An example of a pathology related to nonconvexity is studied in [46], where a CBF-like function was used to identify safe headings of a ship with respect to a stationary object, resulting in a nonconvex set of safe headings. The system in [46] did not admit any continuous safeguarding control law for the ship heading, whereas the proposed discontinuous control law failed to ensure safety in the presence of arbitrarily small disturbances.…”

“…An example of a pathology related to nonconvexity is studied in [46], where a CBF-like function was used to identify safe headings of a ship with respect to a stationary object, resulting in a nonconvex set of safe headings. The system in [46] did not admit any continuous safeguarding control law for the ship heading, whereas the proposed discontinuous control law failed to ensure safety in the presence of arbitrarily small disturbances. The pathology was overcome by using a robust hybrid feedback control law.…”

Four Degree-of-Freedom Hydrodynamic Maneuvering Model of a Small Azipod-Actuated Ship With Application to Onboard Decision Support Systems

“…The second solution is an online method, and uses a virtual autopilot with collision avoidance capabilities to predict and propose a safe ship trajectory. A novel hybrid control barrier function (CBF) formulation is used as basis for the virtual autopilot, building upon some of the authors previous work [21], [22], [23].…”

“…A hybrid CBF design is used for collision avoidance, with target ships assumed to maintain a constant velocity given by the latest AIS information. The algorithm builds upon our previous work on hybrid feedback control applied to obstacle avoidance for underactuated ships [21], safe maneuvering control using nonhybrid CBFs [23], and hybrid CBFs applied to obstacle avoidance of nonholonomic vehicles [22]. We refer to [32] for introduction to CBF theory.…”