1: \begin{abstract}

2: 	In this paper, we propose a model predictive control based two-stage \acl{ems} that aims at increasing the renewable infeed in interconnected \acp{mg}.

3: 	In particular, the proposed approach ensures that each \ac{mg} in the network benefits from power exchange.

4: 	In the first stage, the optimal islanded operational cost of each \ac{mg} is obtained.

5: 	In the second stage, the power exchange is determined such that the operational cost of each \ac{mg} is below the optimal islanded cost from the first stage.

6: 	In this stage, a distributed augmented Lagrangian method is used to solve the optimisation problem and determine the power flow of the network without requiring a central entity.

7: 	This algorithm has faster convergence and same information exchange at each iteration as the dual decomposition algorithm.

8: 	The properties of the algorithm are illustrated in a numerical case study.

9: \end{abstract}

10: