Abstract: A microgrid is a new energy system that integrates distributed generation, energy storage devices and loads. Based on the specific application situation of microgrid, this paper establishes a suitable front-forward dispatching model based on different conditions and considering different factors. These models adopt the big data processing method, and establish the economic evaluation and prediction model of the microgrid by considering the constraints of the battery charging power and the energy exchange between the microgrid and the main grid. Furthermore, the renewable energy supply is introduced as a new variable, and the fixed value constraint is replaced with the range constraint. The goal is to minimize the power supply unit price of the average load, and to establish the optimal day-ahead scheduling model of power width constraint without energy storage. In addition, the charge state and service life of the battery are also considered, and the optimal day-ahead scheduling model considering the energy storage and power exchange constraints is established. For other different conditions, the corresponding models are established, and the comprehensive performance of the optimized scheduling scheme in various aspects is analyzed. Through these models, we can better understand the operation of the microgrid system, and provide useful guidance and decision support for practical applications.