Periodic event-triggered robust trajectory tracking control for underactuated unmanned surface vehicle without velocity measurement.

Abstract

This paper proposes a periodic event-triggered control (PETC) algorithm for underactuated unmanned surface vehicle (USV) subject to unknown exogenous perturbances within limited communication bandwidth. Initially, the underactuated USV model is reformulated using output redefinition-based dynamic inversion (ORDI) so that the system attains a well-defined relative degree through an approximate variable vector. While typical USV control strategies isolate the yaw and surge channel, the ORDI integrates both dimensions into a single framework to adopt a direct controller design. Subsequently, the minimum-learning-parameter radial basis function neural network (RBFNN) with adaptive laws is employed to effectively approximate the nonlinear dynamics and external perturbations with rapid and fewer computations. After that, an anti-chattering velocity observer is presented to provide accurate velocity estimation based solely on position data transmission. Building on this, a PETC algorithm is introduced which balances the periodic sampling with traditional event-triggered control via a sliding mode manifold. This mechanism is designed to assess the requirement for computations and the subsequent transmission of updated measurements alongside current control signals. Furthermore, it can dynamically adjust the communication frequency between the controller and the actuators, in accordance with the digital platform's demands. Moreover, Theoretical analysis rigorously proves that state errors and estimation errors converge to equilibrium and ensure the system stability. Numerical simulations substantiate the robustness and superior performance of the proposed control scheme under bandwidth limitations and uncertain conditions.