Robust predictive control for anti-rolling path following of underactuated surface vessels using adaptive Kalman filter

Guosheng Li,Jun Zhang,Linhe Zheng,Zhilin Liu

doi:10.1177/1729881419877269

Abstract

With the strong time-varying disturbances in wave fields, the large sway motion of underactuated surface vessels can be caused leading to the control performance reduction and large range of roll angle, such that a novel path following control with roll constraints is proposed. Firstly, the affine switching system model is constructed with coupled surge, sway, roll, and yaw. Then, with augmented state adaptive Kalman filter, the estimations of system state and disturbance can be obtained for the feedforward compensation. Considering the existing estimation error between real and estimated states, the state estimation has been adopted directly in the controller design for the performance improvements. With the convex optimization of linear matrix inequalities, the invariant ellipsoid constraint ensures the close-loop system uniformly bounded and stable. Lastly, the simulation results have verified the effectiveness of the designed control approach.

Highlights

Path following of surface vehicles has attracted the attention of the control community for many years, and in the last few years, the motion control problem of typical and asymmetric underactuated surface vessels (USVs) has been a research focus.[1,2,3]
For the path following issue of USVs, the large sway motion and rudder saturation caused by time-varying disturbances can lead to the constraints and robustness problem
Considering the control performance, stability, and safety, a direct constrained predictive control with roll constraints is proposed for path following based on augmented Kalman filtering method

Summary

Introduction

Path following of surface vehicles has attracted the attention of the control community for many years, and in the last few years, the motion control problem of typical and asymmetric underactuated surface vessels (USVs) has been a research focus.[1,2,3] At present, most marine surface vessels are only equipped with the propeller and rudder. For the path following problem of USVs with roll constraint, a robust constrained predictive control method was proposed utilizing adaptive Kalman filter and disturbance observer. To handle the above existing problems, a novel direct constrained robust predictive control combined with augmented state adaptive Kalman filter is proposed for USVs to achieve path following and rudder roll stabilization in this article. (2) Considering the existing estimation error between actual and estimated states, the real measurements are utilized directly in the performance index and control design, such that a direct constrained robust predictive control method combined with adaptive Kalman filter can be proposed for the achievements of the better performance of path following and rudder roll stabilization. V x^ ðk þ 1Þ À V x^ ðkÞ 1⁄4 1⁄2wðkÞx^ ðkÞ þ Bdd1ðkÞ þ KðkÞHeðkÞTP1⁄2wðkÞx^ ðkÞ þ Bdd1ðkÞ þ KðkÞHeðkÞ À x^ TðkÞPx^ ðkÞ

H TK TðkÞP B d w TðkÞP K ðkÞH

H TK T e ðkÞ

Conclusions