AI-Aided Region-Based Active Stabilization in Autonomous DC Microgrids

Creating DC microgrids (MGs) to achieve local source and load balance is becoming a grand possibility with the introduction of inverter-based resources (IBR), e.g. photovoltaics (PV), energy storage systems (ESS), etc. Autonomous DC MGs can be implemented to reduce fossil fuel consumption and energize remote communities. This paper begins with the mathematical modeling of DC-DC source converters, then proceeds to apply a virtual impedance loop to increase autonomous stability of DC MGs. The concept of a stability region boundary (SRB) based on two system parameters is introduced and then determined by the artificial intelligence (AI) aided Kernel Ridge Regression (KRR) approach. The SRB is verified by a four-converter DC MG through time-domain simulation in MATLAB/Simulink.