Fault-Tolerant Model Predictive Control for Safety of Unmanned Surface Vessels Berthing under Multimodal disturbances and Various Constraints

Abstract

Autonomous berthing is a typical task for maritime operations of unmanned surface vessels (USVs). However, during the berthing operations, USVs are subject to multimodal disturbances, such as external ocean disturbances (EODs) and internal thruster faults (ITFs), as well as various constraints, including underactuated nonlinear dynamic constraints, actuator saturation constraints, and obstacle avoidance constraint. In this paper, a fault-tolerant model predictive control (FTMPC) framework is proposed for safety of USV berthing by uniformly considering both disturbances and constraints. Specially, a control density function is integrated into the FTMPC framework to model the obstacle avoidance constraint. Moreover, leveraging the backstepping method and fuzzy logic system, an auxiliary control law considering the EODs and ITFs is constructed as a constraint. Furthermore, sufficient conditions that ensure recursive feasibility, and thus closedloop stability, are provided analytically. Within this FTMPC framework, multimodal disturbances and various constraints can be naturally considered simultaneously. Multiple experimental results in autonomous berthing scenarios demonstrate that the proposed method has excellent fault tolerance and safety under multimodal disturbances and various constraints.