A Wake Avoidance Method for Unmanned Maritime Vehicles using a Trajectory-Circle-Based Control Barrier Function

Efficient obstacle avoidance while controlling unmanned maritime vehicles (UMVs) is an important topic. Evading obstacles that are both time-varying and shapechanging, such as avoiding moving boats and their wake regions that vary with the boat's motion and propagate along time, is essential for UMV's safe operation yet a challenging task. To address this challenge, this paper proposes a novel trajectory-circle-based control barrier function (CBF) method for variable-area obstacle avoidance. As a stepping stone, the boundaries of obstacles are defined by modeling the relationship between the Kelvin wake pattern and the boat's motions, thereby defining the boundaries of obstacles. We propose a novel CBF formulation based on the geometric relationship between the collision area and UMV motion states. Particularly, we leverage the geometrical properties between the predictive circular trajectory and the collision area to construct invariant sets. Compared to traditional CBF-based methods, our design achieves smoother avoidance trajectories and requires smaller control corrections, guaranteeing a safe operation for UMVs.