With the rapid growth of new energy, distributed photovoltaics (DPVs) are accounting for an increasingly high proportion in the power system (EPS) and have become an important component of China's future new EPS. However, the consumption problem of DPV is very prominent, and the problem of abandoned light urgently needs to be solved. Therefore, this article proposes a dynamic optimization operation method for optical storage based on artificial intelligence (AI). Firstly, to address the issue of significant prediction errors in photovoltaic output under different weather conditions, this paper uses attention mechanism and Long Short Term Memory Neural Network (LSTM) to predict DPV output. Then, the peacock optimization algorithm is used to optimize the energy storage configuration in the substation area, plan the optimal site and capacity, set optimization constraints, and deploy optimization nodes. This article constructs a configuration optimization model based on the Peacock Optimization Algorithm, and implements the optimization of energy storage configuration through iterative correction processing. The results show that this strategy can achieve the optimal investment benefits of energy storage, effectively enhance the voltage quality and power stability of the power grid, and enhance the operation level of EPS.
Cheng et al. (Sun,) studied this question.