Abstract: In allusion to the operation of grid-connected photovoltaic (abbr. PV) energy storage microgrid, a two-stage day-ahead energy management strategy containing centralized control and distributed control was proposed. In the first stage, all distributed PV storage units in the microgrid were equalized to one cluster and taking the minimized interaction cost between microgrid and large power grid was as the objective, an economic optimal dispatching model was established and solved by beetle swarm optimization (abbr. BSO) to obtain total discharging power of energy storage units and the power purchasing and selling planning of microgrid. In the second stage, taking the minimized charging loss of energy storage units as the objective a distributed control model was constructed, and by use of incremental cost consensus algorithm utilizing Leader-Follower mode the total energy storage power obtained from the calculation in the first stage was optimally allocated in real time. Finally, the day-ahead economic optimal dispatching of PV energy storage microgrid was realized. The effectiveness of the proposed strategy is verified by the results of analysis on computing example.