0885-8950 (c) 2016 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information. This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TPWRS.2016.2634085, IEEE Transactions on Power Systems 1 Distributed Robust Finite-Time Secondary Voltage and Frequency Control of Islanded Microgrids Nima Mahdian Dehkordi, Nasser Sadati, Member, IEEE and Mohsen Hamzeh, Member, IEEE Abstract—This paper presents a distributed, robust, finite-time secondary control for both voltage and frequency restoration of an islanded microgrid with droop-controlled inverter-based distributed generators (DGs). The distributed cooperative sec- ondary control is fully distributed (i.e., uses only the information of neighboring DGs that can communicate with one another through a sparse communication network). In contrast to existing distributed methods that require a detailed model of the system (such as line impedances, loads, other DG units parameters, and even the microgrid configuration, which are practically unknown), the proposed protocols are synthesized by considering the unmodeled dynamics, unknown disturbances and uncertain- ties in their models. The other novel idea in this paper is that the consensus-based distributed controllers restore the islanded microgrid’s voltage magnitudes and frequency to their reference values for all DGs within finite time, irrespective of paramet- ric uncertainties, unmodeled dynamics, and disturbances, while providing accurate real power sharing. Moreover, the proposed method considers the coupling between the frequency and volt- age of the islanded microgrid. Unlike conventional distributed controllers, the proposed approach quickly reaches consensus and exhibits a more accurate robust performance. Finally, we verify the proposed control strategy’s performance using the MATLAB/SimPowerSystems toolbox. Index Terms—Distributed cooperative secondary control, feed- back linearization, finite-time control, microgrids, multi-agent systems, power sharing, robust control. I. I NTRODUCTION Microgrids as modern, small-scale conventional power sys- tems consist of renewable energy sources, such as wind gener- ators (WTs), photovoltaics (PVs), and micro turbines (MTs); energy storage systems; and local loads that can operate in both grid-connected and islanded operating modes. Microgrids offer more efficiency and reliability than conventional power grids. In the grid-connected mode, while the microgrid is connected to the main grid, the main grid determines the microgrid’s voltage and frequency. In this case, the microgrid delivers the preplanned scheduling real and reactive powers to the main grid. In the islanded mode operation, unpredictable disturbances or preplanned scheduling cause disconnection of the microgrid from the main grid. As a result, the islanded microgrid’s pre-islanding control strategy can make it unstable. To maintain the voltage and frequency stability of dis- tributed generators (DGs) in the microgrid, an effective control strategy, called primary, should be employed [1], [2]. However, because primary control has some drawbacks, such as the N. Mahdian and N. Sadati are with the Department of Electrical Engi- neering, Sharif University of Technology, Tehran 113659363, Iran (e-mail: mahdian dehkordi@ee.sharif.edu; sadati@sharif.edu). M. Hamzeh is with the Department of Electrical Engineering, Shahid Beheshti University, Tehran 1983963113, Iran (e-mail: mo hamzeh@sbu.ac.ir). voltage and frequency deviation caused by droop technique, hierarchical control has been presented to standardize the mi- crogrid operation [3]–[6]. Hierarchical control consists of three layers: primary, secondary, and tertiary control. The secondary control compensates for the primary control’s deviation, while the tertiary control is responsible for economic dispatch and power flow optimization issues. This paper focuses on the secondary control of an islanded microgrid, presenting a distributed robust finite-time control structure for an islanded microgrid system. The traditional secondary control uses a centralized structure that requires all the information of the individual DGs and a central com- puting and communication unit. These requirements reduce the overall system reliability and increase its sensitivity to failures that can lead to a single point of failure [4], [5], [7]. To overcome the aforementioned drawbacks, several recent studies have proposed the distributed secondary control as a promising approach [8]–[21]. The distributed cooperative secondary control uses only the information of neighboring DGs that can communicate with one another through a sparse communication network. The distributed secondary control has the advantages over a central control structure of increasing system reliability, decreasing its sensitivity to failure, and elim- inating the need for a central computing and communication unit. Moreover, because of plug-and-play capability of micro- grids, microgrid’s physical and communication structures can be time-varying. The distributed control structure provides a robust secondary control framework that properly works irrespective of time-varying, restricted, and unreliable com- munication networks. Centralized control structure requires a complex communication network with two-way communica- tion link [7], [10], [11], [17]–[20]. This requirement reduces the overall system reliability and increases its sensitivity to failures that can lead to a single point of failure. However, the distributed controllers communication network is sparse, and each agent communicates with local neighbors. Unlike a fully connected network, this sparse network reduces the communication infrastructure costs and makes it scaleable and reliable. Because more extensions increase the controller com- plexity, scalability of central controllers is not straightforward. Moreover, the failure of any unit can shut down the whole system. In summary, a distributed control strategy has the advantages of surviving uncertainty and disturbances. It also improves plug-and-play capability, security, and reliability. Moreover, it allows you to have easier scalability, simpler communication network, and fully distributed data updating leading to efficient information sharing, and finally to make faster decisions and operations [14]–[21].