A two-stage energy management for plug-in hybrid tugboats based on MPC method

Hybrid configuration for ships holds the potential to significantly enhance energy efficiency and reduce CO2 emissions which are already used on several ship types. This configuration presents an additional challenge in determining the optimal power distribution among multiple energy sources for more efficient, cost-effectiveness, and clean energy utilization. Currently, the rule-based strategy is widely used, which has poor flexibility. In order to further improve the performance, it is necessary to investigate better control strategies. In this work, a two-stage energy management strategy based on Mixed Integer Nonlinear Programming (MINLP) and Model Predictive Control (MPC) was proposed for plug-in hybrid tugboats equipped with diesel generators and energy storage devices, to achieve real-time optimal power distribution of multiple energy sources during voyages. Based on the actual operation data of the tugboat, a typical tugboat load curve was generated. Aiming at minimizing the equivalent fuel consumption, a mixed integer nonlinear optimization model was established, and the adjustable battery equivalent fuel factor and a series of constraints were considered. By solving the global optimal problem in the predicted time domain, the reference tracks of MPC were obtained, and finally the real-time optimal energy scheduling was realized. Simulation results demonstrate that the real-time optimization-based MPC (RT-MPC) energy management strategy can reach similar optimization effects compared with the global optimization-based MPC (G-MPC) energy management strategy, and the deviation is 0.726%. In addition, through the robustness verification, RT-MPC strategy can effectively deal with the load fluctuations and realize the real-time optimization of the system