Article Dans Une Revue International Journal of Electrical Power & Energy Systems Année : 2024

An energy trade-off management strategy for hybrid ships based on event-triggered model predictive control

Résumé

This paper addresses the energy management problem of hybrid ships by proposing an event-triggered model predictive control (ET-MPC) method. The novelty in this work lies in the establishment of an event-triggered mechanism and a state prediction model for energy management of hybrid ships. First, torque models of the internal combustion engine (ICE) and electric machine (EM) are developed using a data-driven approach, followed by the construction of fuel consumption and carbon emission models. Second, an event-triggered mechanism, dependent on state prediction error, is introduced and updated at each time step based on the system's current state. Additionally, a cubature Kalman filter (CKF) is employed to estimate and correct the state prediction error, minimizing inaccuracies. A trade-off coefficient is incorporated to optimize the balance between fuel consumption and carbon emissions. The ET-MPC method results in a 0.68% difference in fuel consumption and 3.43% increase emissions compared to the traditional MPC method. However, ET-MPC significantly reduces computational overhead by 56.66. The ET-MPC method effectively allocates the ship's energy according to the varying trade-off coefficient, achieving optimal energy management under different constraints.

Fichier principal
Vignette du fichier
1-s2.0-S0142061524005350-main.pdf (2.38 Mo) Télécharger le fichier
Origine Publication financée par une institution
Licence

Dates et versions

hal-05078652 , version 1 (22-05-2025)

Licence

Identifiants

Citer

Diju Gao, Long Chen, Yide Wang. An energy trade-off management strategy for hybrid ships based on event-triggered model predictive control. International Journal of Electrical Power & Energy Systems, 2024, 162, pp.110312. ⟨10.1016/j.ijepes.2024.110312⟩. ⟨hal-05078652⟩
49 Consultations
67 Téléchargements

Altmetric

Partager

  • More