This paper addresses a new decentralized framework for optimal operation of multi-microgrids (MMGs) in both normal operation and self-healing modes. For secure scheduling, two reliability-oriented indices are considered in self-healing mode. When a fault happens in a MG, the operation changes to the self-healing mode and the on-fault MG receives power support from other normal-operated MGs. The power support shares are calculated according to the reliability-oriented indices. The main idea is to equalize the power support shares with power exchanges among the MGs. In both normal and selfhealing modes, the objective functions are to schedule the distributed generators (DGs) and the power exchanged between each MG for minimizing the operation costs. The proposed model is effectively linearized and implemented on a test MMGs system. The simulation is run over a 24-hour scheduling time horizon. Analyzing the results over some faulty scenarios verifies the impressiveness of the presented model.
