Abstract: In order to solve the bottleneck problem of high life loss cost of energy storage in microgrid, a day-ahead and intra-day two-stage coordinated scheduling model and control strategy based on double battery group are proposed. In the process of day-ahead scheduling, an economic dispatch model with the minimum total cost of microgrid operation is built, which considers daily forecast data, life of energy storage and market price, and a random weight particle swarm optimization method is used to calculate the function. In the process of intra-day scheduling, an energy storage system based on the topology of double battery group is built to smooth tie-line power fluctuation caused by the forecast errors of intermittent distributed generation. A real-time control strategy is designed based on the battery characteristics, in which two battery groups are used as a discharging group and a charging group respectively to compensate positive and negative errors of tie-line power. If any battery group arrive its upper or lower limit of state of charge, the two battery groups should switch their dis-charging/charging status simultaneously. Finally, taking a demonstration project of microgrid in a park as an example, the experimental results verify the effectiveness of the day-ahead model and algorithm, and also prove that the control strategy proposed in this paper can effectively prolong the cycle life of battery and improve systematic economy.