Neural Quantized Obstacle Avoidance Control for Underactuated USVs based on Control Barrier Functions

Safety-critical control has received extensive attention in mobile robotic systems such as unmanned surface vehicles (USVs), whose main objective is to make the vessels accomplish their desired tasks while avoiding unsafe areas, also known as obstacle avoidance control. In this paper, for a class of underactuated USV systems with collision avoidance requirements, a neural quantized obstacle avoidance tracking control algorithm based on control barrier functions (CBFs) is developed, which is implemented by backstepping technique and can guarantee the semi-globally asymptotically stable of the whole closed-loop system subject to the safety conditions. Moreover, by introducing RBF neural networks and quantizers, the proposed control strategy allows the USV systems to have unknown parameters, unknown nonlinearities and input quantization concurrently. Simulation results demonstrate the efficiency of the presented approach.