Abstarct: The problem of stability in power systems has been studied since 1970s. The increasing penetration of distributed generations and in the near future in the form of microgrids, can face the power system to some new challenges. Most of distributed generations are lacking the rotating inertia. They are connected to the grid by power electronic converters which their dynamics are determined by their control algorithms. So it seems necessary to investigate power system stability in presence of distributed generation sources. Microgrids will be effective on dynamics of the system if their penetration level is increased and there would have existed hundreds or even thousands of active micro-grid cells. This is the situation in which multi-micro-grids is defined. Because of complexities in studies of power system dynamic, network reduction and dynamic equivalent methods should be used to ease and reduce the computation time. On the other hand, because of different structure and system, conventional dynamic equivalent methods are not useful in active distribution networks. Therefore, it has been suggested to use identification methods. In the present thesis, after investigating the mentioned dynamic equivalent methods for active distribution networks, a suitable model is chosen. Then the role of microgrids in small and large signals stability of power system is studied. The role of passive and active microgrid cells in power system stability is also addressed for better evaluation. It can be seen that the passive cell reduces system damping and may lead the system to instability. Using the distributed generation sources that are baxsed on power electronics, as a part of an active distribution network cell, it's possible to provide a certain level of damping and maintain the system stability by adjusting active and reactive power of these sources.