Real-time power management of integrated power systems in all electric ships leveraging multi time scale property

Abstract—All-electric ships (AES), enabled by integrated power systems (IPS), have been pursued for both commercial and military applications to meet the increasing ship-board power demand and environmental sustainability initiatives. They necessitate real-time power management (PM) for dynamic reconfiguration to support system critical operations in the event of dynamic load change or IPS component failures. The nonlinear, large scale trajectory opti-mization problem associated with IPS, along with the non-analyt-ical nature of IPS model, makes many existing methods inadequate in meeting the real-time requirements. In this paper, we develop a methodology that exploits time scale separation, a characteristic associated with IPS dynamics, to achieve real-time optimization. In parallel, a dynamic model of the IPS with gas turbine and fuel cell as power plants is developed that captures the relevant dynamics but is simple enough for real-time optimization. The tradeoffs be-tween the computational efficiency and optimization accuracy are analyzed. The optimization results for IPS PM on a real-time sim-ulator are reported, which illustrate the real-time feasibility of the proposed optimization strategy. Index Terms—All-electric ships (AES), integrated power systems (IPS), optimal power management, multitime scale property, real-time large scale optimization. I