Model-based two-layer control design for optimal power management in wind-battery microgrids

Authors: Pablo R. Baldivieso-Monasterios, George C. Konstantopoulos, Antonio T. Alexandridis

Published in:  Journal of Energy Storage, Volume 48, April 2022, 104005 DOI: https://doi.org/10.1016/j.est.2022.104005

Date Published: 21 Jan 2022

Abstract: 

In this paper, a comprehensive model in Hamiltonian form of a Microgrid (MG) composed of heterogeneous components, i.e. wind turbine generator, battery storage and local loads together with their power conversion units, is developed. The proposed model analytically captures the energy conversion capabilities of different sustainable energy sources. Based on this model description, novel primary (nonlinear PI) and secondary controllers (receding horizon) are proposed that ensure boundedness of the currents injected by each energy source and optimal power management operation of the entire MG. Furthermore, closed-loop stability analysis is rigorously proven for both primary and secondary control loops taking into account the accurate Hamiltonian description of the whole MG that includes the energy conversion characteristics. Detailed simulation results of the entire MG connected to a weak grid and operating in islanded mode are provided to validate the proposed model, the control design and the stability analysis under various scenarios.

Keywords: Microgrids; Hamiltonian modelling; Primary and secondary control; Stability analysis

Insights for EnergyREV:

In this paper, we model a Microgrid from generation (renewable input and storage) to grid connection using a port-Hamiltonian approach. We study the effect of shifting classical droop characteristics to the cost of an optimal controller. The benefits of such approach are twofold: removing the slow dynamic response of converter-based renewable sources and allowing for the inclusion of more performance criteria. We prove the stability and feasibility of the optimal control problem. We test our approach in Microgrids containing nonlinear loads, and  weak grids.