Composite Controller Design for AUV With Uncertainties and Noises Based on Combined Kalman Filter and RBFO

This article investigates the veloctiy control problem of autonomous underwater vehicle(AUV) with disturbances and measurement noises. First, the disturbances, nonlinearities and uncertainties are regard as lumped disturbances. Second, a special combination of Kalman filter(KF) and Radial Basis Function Neural Network Observer(RBFO) is proposed to filter noises and estimate lumped disturbances. And then, under the premise of ensuring that multiple observers maintain an equivalent level of noise amplification, their capabilities in reconstructing system states have undergone a meticulous comparative assessment. Furthermore, the tracking capability of the composite controller based on the KFRBFO for velocity is verified. Through qualitative comparison, the advantages of the proposed method in terms of state reconstruction and control performance over existing methods are revealed.