Adaptive distributed secondary control of microgrids via single‐network adaptive dynamic programming method.

Summary: In the presence of parametric uncertainties for microgrids, this paper proposes an adaptive secondary control method via the distributed adaptive dynamic programming theory. For the controller design, a single critic neural network is introduced to deal with the uncertainties in the model of distributed generators. By using such technique, the secondary controller can be trained directly from data measurements and thus can adapt to system or external disturbances. In addition, the proposed controller is distributed, which means that the realization of the controller only needs the information of each distributed generator and its neighbors on the communication digraph. Thus, the requirement for a central controller is obviated and the system reliability is improved. Accordingly, stability of the closed‐loop system is proved by Lyapunov technique. Finally, 4 scenarios including the islanding process, changes in system parameters, load variation, and changes in communication topology are given to demonstrate the effectiveness of the proposed method by a microgrid test system in MATLAB/SimPowerSystems. [ABSTRACT FROM AUTHOR]