Multi-Port DC Microgrids: Online Parameter Adaptation in Model Predictive Control

This paper investigates control of multi-port DC microgrids with a fast transient response and online parameter adaptation. This DC microgrid can handle connection and disconnection of several energy generation units (e.g., fuel cell or photovoltaic units) and battery energy storage units to provide a DC load. To control several unidirectional and bidirectional converters of this microgrid, a scalable model predictive control (MPC) with online parameter adaptation capability is proposed. This adaptive model predictive control (AMPC), which provides fast MPC control for both inner and outer control loops, obviates the need to have a good knowledge about system parameters (as required by conventional MPC controllers). The estimated values of system parameters are deployed in the MPC controllers to ensure fast transient response of the system. This controller increases the scalability of the system to add or remove new power units with unknown parameters. Simulation case studies validate the proposed controller structure.