Dynamic STA-based quantized trajectory tracking control for unmanned surface vehicles without velocity measurements

To address the trajectory tracking problem of underactuated unmanned surface vehicles (USVs) subject to external disturbances and communication constraints caused by limited transmission resources, this paper proposes an improved dynamic super-twisting-algorithm (IDSTA)-based quantized control strategy without velocity measurements.

First, considering that the underactuated USV system lacks a relative degree in the input-output channel, a virtual input-based dynamic inversion method is introduced to establish the required relative degree. Subsequently, a second-order observer is developed to accurately estimate both trajectory and velocity states in the absence of direct velocity measurements. In addition, a radial basis function neural network (RBFNN) is employed to online approximate unknown nonlinear dynamics. Based on these components, an improved dynamic STA controller is designed to suppress chattering and enhance system robustness while explicitly incorporating input quantization effects into the control framework.

The simulation results demonstrate that the proposed velocity observer can estimate the actual velocity within finite time. Furthermore, the proposed control strategy achieves high-precision trajectory tracking while maintaining excellent transient and steady-state performance in complex marine environments. Specifically, the position tracking errors converge to within 0.2 meters and the yaw-angle tracking errors are constrained within 0.01 rad. Meanwhile, the communication frequencies in the surge and yaw channels are reduced by 89% and 72.9%, respectively.

The proposed control algorithm provides accurate trajectory tracking performance for underactuated USVs through the integration of a velocity observer, a RBFNN, and an improved dynamic STA. The research findings provide a valuable reference for the reliable control of the USVs operating in the complex marine environments.