Robust optimisation-based order reduction and stability analysis of autonomous DC microgrid with consideration of non-linearity

Abstract

The reduced-order modelling in the context of timescale separation of DC microgrid based on discrimination of fast and slow states is an important concept that is implemented in this paper. DC microgrid composing of fuel cell, which has inherent sluggish response than that of other components, such as PV and battery, which has faster response, forms an important basis for timescale separation. Further, the states associated with high non-linearity that does exist in fuel cell are retained intact while the rest of the electrical components have been reduced by robust optimization through mixed H2/H∞ norm by Particle Swarm Optimization (PSO) algorithm. The feasibility of nonlinear stability analysis, on the contrary to small signal analysis of the reduced-order system, is addressed by Input to State stability (ISS) criteria, which guarantee Bounded Input Bounded State (BIBS) stability. Series of simulation results and different theorems addresses the objective implemented. © 2019 John Wiley & Sons, Ltd.