Linear State-Feedback Primary Control for Enhanced Dynamic Response of AC Microgrids

• Published in: IEEE transactions on smart grid Vol. 10; no. 3; pp. 3149 - 3161
• Main Authors: Perez-Ibacache, Ricardo, Silva, Cesar A., Yazdani, Amirnaser
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 01.05.2019; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Actuation; Complex oscillator network; Computer simulation; Control systems; decentralized control; Design optimization; distributed energy resource; Distributed generation; Dynamic response; Electric potential; Electric power grids; Energy resources; Energy sources; islanded and grid-connected mode of operation; linear quadratic estimator and regulator; Linear quadratic Gaussian control; Microgrids; Multivariable control; Observers; Oscillators; Power system stability; primary control of microgrids; Robustness; Robustness (mathematics); State feedback; State observers; State variable; Time domain analysis; Trajectory optimization; Voltage control
• ISSN: 1949-3053; 1949-3061
• DOI: 10.1109/TSG.2018.2818624

This paper proposes a state feedback primary control strategy for microgrids with multiple distributed energy resource units, improving their transient behavior in both islanded and grid-connected modes of operation. To that end, the interaction of each distributed energy resource unit within the microgrid is modeled as a lumped dynamic system, which results to be nonlinear and multivariable. For the closed-loop control of such multivariable systems, the full state feedback formulation is preferred which requires a suitable state observer. For the design of the observer and feedback gains, the solution of the linear-quadratic estimation and regulation problems is considered. For simplicity, an approximate linear model at a representative operating point is derived. The linear quadratic Gaussian/loop-transfer recovery is adopted as the design procedure to optimize the trajectory of the state variables subject to a desirable actuation effort, in this case of the voltage amplitude and frequency, yielding a solution that is robust to model uncertainties. The effectiveness of the strategy is assessed through time-domain simulation on the CIGRE benchmark medium-voltage distribution network with three distributed energy resource units. These results are compared to those obtained with conventional static droop gains and with a state-of-the-art technique from the literature.