High-speed multihull anti-pitching control based on heave velocity and pitch angular velocity estimation

Abstract

This paper proposes a novel anti-pitching control algorithm based on algebraic model predictive control (AMPC) for high-speed multihull, in which the heave velocity and pitch angular velocity cannot be measured directly. Specifically, a multihull vertical control model is established with the employed anti-pitching appendages, and the uncertainty of the model as well as the coupling between heave and pitch motion are investigated. To address unmeasurable of the heave velocity and pitch angular velocity, a novel kinematics-based Kalman filter is designed to estimate these states online, which is substantially different from the existing works. Then, the AMPC strategy for varying receding-horizon optimization is proposed, which significantly reduces the amount of online calculation. To estimate the lumped uncertainty, the second-order filter data of input and output is used to design a disturbance observer with less parameters, which can perform feed-forward compensation to enhance the robustness. The convergence of the disturbance observer and the closed-loop system is analyzed mathematically. Finally, the advantages of the proposed anti-pitching control approach are demonstrated in both theory and simulation.