Objective In order to balance the relationship between multiple objectives in the photovoltaic-energy storage DC distribution network and realize the optimal scheduling of energy storage capacity, an optimal scheduling method for energy storage capacity of the photovoltaic-energy storage DC distribution network considering the uncertainty of large-scale access load is proposed.

Methods Firstly, the uncertainty of large-scale access load was deeply studied, and a mathematical model of load uncertainty was established to provide a basis for the subsequent optimal scheduling of energy storage capacity. Considering this uncertainty factor, an optimal scheduling model for energy storage capacity of the photovoltaic-energy storage DC distribution network was constructed, whose objective functions were minimizing the expected difference between photovoltaic output and load and minimizing the daily operating cost, with corresponding constraint conditions established respectively. The multi-objective grey wolf optimizer (MOGWO) was used to optimize the constructed objective function under the condition of satisfying the constraints, and the obtained optimal solution was the optimal scheduling of energy storage capacity for the photovoltaic-energy storage DC distribution network. The performance of the proposed algorithm was compared with that of the commonly used particle swarm optimization (PSO) algorithm and random forest (RF) algorithm in the optimal scheduling of energy storage capacity of the photovoltaic-energy storage DC distribution network.

Result Experimental verification shows that compared with traditional algorithms, the proposed method can optimize the utilization efficiency of photovoltaic energy, effectively reduce energy loss through improved strategies, and thus significantly improve the overall energy utilization efficiency. In addition, under different light intensities, the operation stability of the distribution network can be higher than 88%; even under different weather conditions, the proposed method can still achieve efficient and stable scheduling.

Conclusion It is proved that the proposed method can flexibly adjust the charge-discharge strategy of energy storage equipment according to actual operation needs, realize the efficient utilization of photovoltaic energy and the optimal allocation of energy storage capacity, and provide a new idea for the optimal operation of the photovoltaic-energy storage DC distribution network.