Prescribed Performance-Based Collision Avoidance and Path-Following for Underactuated Surface Vehicle

Abstract

This paper presents a novel prescribed-time path-following control framework for underactuated surface vehicles (USVs) operating in cluttered environments with limited sensing and communication resources. Unlike existing asymptotic or finite-time methods requiring velocity measurements, the proposed approach guarantees prescribed-time convergence in a velocity-free manner. A smooth finite-range potential function is integrated with a flexible prescribed-performance mechanism to ensure safe obstacle avoidance without violating tracking constraints. A nonlinear prescribed-time extended state observer (NPTESO) is developed to estimate unmeasured velocities and disturbances with convergence time independent of initial conditions. Furthermore, hybrid input quantization effects are explicitly addressed via adaptive compensation, ensuring robustness under limited velocity sensing and quantized actuation. Numerical simulations demonstrate fast convergence, improved tracking accuracy, and reduced control effort in obstacle-populated environments.