Event-triggered predictive path following control of autonomous ships with an MMG model

Abstract

To improve the energy efficiency and path following performance of autonomous ships, this paper proposes a Maneuvering Modeling Group (MMG) model based Event-Triggered Model Predictive Control (MET-MPC) method. Firstly, the path following control is transformed into a ship course control problem with the improved Line of Sight (LOS) guidance law considering drift angle, and an MMG model is applied to deal with variable sailing speeds. Then, a hyperbolic tangent function is introduced as the triggering threshold to reduce computational burden and energy consumption. Specifically, a dynamic buffer is introduced to store the optimal control sequence computed during each event triggering, which can keep the previous control input with no event triggered. Simulation experiments show that, the proposed MET-MPC method has better energy-saving performance and can reduce the steering frequency by up to 38.1% compared to the method without using the event-triggered mechanism.