Optimization Configuration Method of Energy Storage Considering Photovoltaic Power Consumption and Source-Load Uncertainty

The proposal of a "double carbon" target has resulted in a gradual and continuous increase in the proportion of photovoltaic (PV) access to the distribution network area. To enhance the capability of PV consumption and mitigate the voltage overrun issue stemming from the substantial PV access proportion, this paper presents a multi-objective energy storage optimization allocation methodology. Firstly, a PV and load uncertainty model is established based on Beta and Normal distributions, and the Monte Carlo Method (MC) is used to simulate the annual output. The representative scenarios are obtained after the combination is cut down by the K-means clustering algorithm. The poor operation scenarios are identified by using the variation of the maximum total node voltage deviation. Finally, an energy storage optimization allocation is proposed. Subsequently, the objective function, which seeks to minimize the total daily operating cost of the energy storage system and the PV abandonment rate, is constructed using the evaluation-based function method. The constraints, including node voltage and energy storage device, are considered, and the model is solved using the improved grey wolf optimization algorithm. Finally, a distribution network station system is employed to compare and analyze the capacity allocation results of poor operation scenarios under different schemes, the results showed that after energy storage optimization, the voltage of each scheme decreased from 0.4869kV to 0.42kV, and the curtailment rate decreased by 13.09%, 21.33%, and 23.69% respectively compared to before optimization,which verifies the effectiveness of the models and methods proposed in this paper.