Study on energy management of dual-diesel generator sets hybrid power ships based on model predictive control

Objectives The marine diesel-electric hybrid system reasonably distributes the output power of the diesel engine and motor, which can significantly reduce fuel consumption and emissions. Aiming at the contradiction between the optimal performance and real-time operation of traditional energy management strategies applied to hybrid power systems, this study proposes to implement model predictive control (MPC) to achieve the instantaneous optimization of energy management.

Methods First, an energy flow model of a passenger-ferry hybrid system consisting of dual diesel generator sets, energy storage systems and shore power is established by employing the reverse modeling method. An MPC energy management algorithm that can be solved online by rolling optimization under system constraints is then proposed, taking the total greenhouse gas (GHG) emissions of fuel consumption and electric energy consumption as the objective function. Finally, the sensitivity analysis of the variable prediction horizon lengths is carried out.

Results The simulation results show that the MPC method can reduce fuel consumption by 4.85% and total carbon dioxide emissions by 3.54% respectively compared with the traditional rule-based control method.

Conclusions The MPC method achieves lower fuel consumption, lower carbon emissions and lower computing load than the traditional rule-based control method, giving it promising potential for real ship application.