Abstract: In order to improve the safety and stability of the power grid, this study proposes a collaborative optimization scheduling method for distributed energy storage systems based on the deep Q network model. In response to the dispersed scale and dynamic coupling characteristics of distributed energy storage systems, this study takes multiple data such as photovoltaic output, load demand, energy storage state of charge, and time of use electricity price as state inputs, and uses DQN"s deep neural network to approximate the action value function, thereby achieving dynamic optimization of charging and discharging strategies. This study adopted the DQN algorithm and effectively alleviated the overfitting problem during the training process by introducing the target network and experience replay mechanism. Furthermore, the hyperparameters of the neural network were optimized, thereby improving the convergence speed of the model. The experimental results show that the collaborative optimization scheduling method for distributed energy storage systems based on deep Q-network models not only significantly reduces the operating costs of the system, but also greatly enhances the grid"s ability to absorb renewable energy.