In order to ensure the reliability and high efficiency of the optimal scheduling strategy of distributed energy system, this paper combines big data technology to study the energy storage system of distributed energy system. In order to simplify the daily scheduling optimization of regional DES, the loss of energy storage device and energy conversion device is ignored in this paper, and the output power, conversion power and interworking power of specific equipment are treated linearly in the energy supply model to a certain extent, so as to better simulate the scheduling process. Moreover, this paper combines the big data algorithm to avoid the algorithm falling into local optimum by improving the group learning method. Taking the distribution system as the reference system, the electric energy demand is met by the power grid, the heat energy demand is generated by the gas boiler, and the cold energy is generated by the electric refrigeration. In addition, in different seasons, the comprehensive benefits of the energy supply model in this paper are compared and analyzed. The research results show that the method in this paper has certain effect.