I. Introduction
Global climate change and ever increased energy demand have resulted in increasing integration of distributed energy resource (DER) units and emergence of the microgrid concept [1]. This integration is changing the conventional passive distribution network to the modern active ones. Conventionally, the overcurrent relays are employed to protect the distribution network. The operating principle of these relays is that the magnitude of fault current supplied by the main grid is several orders higher than that of normal line current. In the case of grid-connected microgrids, there is no problem for the detection of faults due to the high short-circuit capacity of the main grid. In the case of standalone microgrids, where the grid does not contribute to the fault current, there are two conditions, depending on the DER interface type. If DER units are interfaced to the microgrid through rotating machines, overcurrent relays can properly operate due to large fault current supplied by these machines. However, fault detection in inverter-interfaced standalone microgrids is a serious challenge [2]. It is due to this fact that the low thermal inertia of voltage-sourced converters (VSCs) makes it necessary to protect the semiconductor switches against the large currents. To this end, VSC current is limited to two times the rated current, commonly using a current limiting strategy embedded in the VSC control system. On the other hand, another protection challenge arisen in both grid-connected and standalone microgrids is bidirectional power flow in the presence of DER units. Conventional overcurrent relays are set based on unidirectional power flow from the main grid to the loads and proper selectivity is achieved using the inverse time-current characteristic curve. Thus, low fault current and bidirectional power flow make the conventional overcurrent-based protection ineffective for active distribution networks.