Frequency-adaptive wave disturbance observation and prescribed performance fast super-twisting control for ROV station-keeping

Abstract

This paper addresses the station-keeping control problem of remotely operated vehicles (ROVs) under near-surface head-sea conditions, where narrow-band wave disturbances induce significant periodic motion. Instead of treating wave effects as unstructured lumped disturbances, a frequency-adaptive disturbance rejection framework is proposed by combining EKF-based frequency estimation, harmonic extended state observation, and prescribed-performance fast super-twisting sliding mode control. The EKF estimates the equivalent dominant disturbance frequency from the measured heave velocity, without requiring direct free-surface wave-elevation measurements. The estimated frequency is then embedded into the HESO internal model to improve the phase-aligned reconstruction of wave-induced disturbances. Based on the estimated disturbances, an initial-condition-independent PPC-FSTSMC is developed to constrain tracking errors and enhance robustness against residual disturbances and model mismatch. Simulation results under wave disturbance, measurement noise, generalized actuator nonidealities, and hydrodynamic parameter bias demonstrate that the proposed method improves station-keeping accuracy and reduces cumulative control effort and input energy relative to the compared controllers.