This paper presents a novel approach to conceive the secondary control in droop-controlled microgrids (MGs). The conventional approach is based on restoring the frequency and amplitude deviations produced by the local droop controllers by using an MG central controller (MGCC). A distributed networked control system is used in order to implement a distributed secondary control (DSC), thus avoiding its implementation in MGCC. The proposed approach is not only able to restore frequency and voltage of the MG but also ensures reactive power sharing. The distributed secondary control does not rely on a central control, so that the failure of a single unit will not produce the fail down of the whole system. Experimental results are presented to show the feasibility of the DSC. The time latency and data drop-out limits of the communication systems are studied as well.
