A thrust allocation strategy for intelligent ships based on model prediction control

Abstract

In order to solve the problem of the traditional sequential quadratic programming thrust allocation method, such as limited feasible region and low propulsion efficiency, a thrust allocation strategy based on model predictive control is proposed in this paper. The proposed thrust allocation strategy is realized by rolling optimization of model predictive control, and moreover, considering the possibility of thruster failure, a fault-tolerant thrust allocation strategy is proposed. The simulation results show that the proposed strategy can not only make the ship run in the expected condition but also reduce the average power consumption by 16.3% compared with the traditional sequential quadratic programming method and has good fault tolerance in the face of thruster failure.

Design of Constraints for a Neural Network based Thrust Allocator for Dynamic Ship Positioning

“…Model Predictive Control (MPC) methods have also been tried to solve TA problem. Zhu et al [5], develops a TA scheme with fault tolerance using an MPC method. A decoupled TA scheme for inland vessels with a feedback mechanism from TA to controller is achieved by Billet et al in [6] using an MPC formulation.…”