An enhanced predictive hierarchical power management framework for islanded microgrids

Jimiao Zhang, Jie Li, Ning Wang, Ben Wu
2021 IET Generation, Transmission & Distribution  
This paper proposes an enhanced three-layer predictive hierarchical power management framework for secure and economic operation of islanded microgrids. The tertiary control, guaranteeing the microgrid economic operation, is built upon the semi-definite programming-based AC optimal power flow model, which periodically sends power references to secondary control. To mitigate uncertainties arising from renewable generations and loads, a centralized linear model predictive control (MPC) controller
more » ... is proposed and implemented for secondary control. The MPC controller can effectively regulate the microgrid system frequency by closely tracking reference signals from the tertiary controller with low computational complexity. Droop-based primary controllers are implemented to coordinate with the secondary MPC controller to balance the system in real time. Both synchronous generators (SGs) and solar photovoltaics (PVs) are simulated in the microgrid power management framework. A unified linear input-state estimator (ULISE) is proposed for SG state variable estimation and control anomaly detection due to compromised cyberphysical system components, etc. Simulation results demonstrated that SG states can be accurately estimated, while inconsistency in control signals can be effectively detected for an enhanced MPC. Furthermore, comparing with conventional proportional-integral (PI) control, the proposed hierarchical power management scheme exhibits superior frequency regulation capability whilst maintaining lower system operating costs. INTRODUCTION The microgrid hierarchical control [1-2] has been studied for a decade, inspired by the bulk power systems operation framework. The system control architecture is commonly divided into three layers: primary, secondary, and tertiary. The primary control takes care of instantaneous system fluctuations due to loads and renewable energy sources (RESs) and allows for a realtime power balancing and sharing among distributed generators (DGs). Primary control commonly relies on droop control. The secondary control eliminates the steady-state frequency and voltage deviations caused by real-time primary control [3] . The tertiary control is responsible for economic dispatch (ED) and reactive power control to manage power flows. Two methods are commonly studied in the literature for implementing different control layers: centralized [4,5] and distributed [6] [7] [8] . The distributed control has well recognized advantages, including This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.
doi:10.1049/gtd2.12297 fatcat:2t2qlplb7re27hj4rob6ybjaqa