Holomorphic embedding power flow modeling of autonomous AC/DC hybrid microgrids

Developing an effective and reliable power flow (PF) tool is of great significance for the steady-state analysis of autonomous AC/DC hybrid microgrids (MGs). In this paper, a generalized PF model is constructed based on holomorphic embedding (HE). With a well-designed embedding technique, the nonlinear PF problem is modeled into an embedded system, which comprehensively accounts for the distinct features of islanded hybrid MGs (such as droop-regulated distributed generations (DGs) and AC/DC interlinking converters (ICs), the variation of system frequency and its effect on admittance matrix and loads in AC part). The embedded system has a generalized structure with a constant sparse matrix, which not only enables it to be solved recursively and efficiently, but also facilitates wide applicability and convenient operation. Moreover, the proposed HE-based model features a deterministic germ that relates to a physical state, which allows it to inherit the merit of the canonical embedding in converging to upper-branch solution reliably and unambiguously, without dependency on initial estimates of state variables. The feasibility and applicability of the proposed model are validated on 3 test systems of different sizes. Furthermore, the computational efficiency and convergence performance are also evaluated.