基于光伏功率预测的储能容量确定研究

Research on Determining Energy Storage Capacity Based on Photovoltaic Power Forecast

针对光储电站建设初期采集装置尚未完善所导致的数据缺失问题,采用Elman前馈神经网络预测光伏功率曲线,保障了数据的可靠性和完整性。此外利用Boltzmann函数拟合了蓄电池相对容量随环境温度的变化曲线,考虑环境温度变化对储能蓄电池相对容量的影响,确保储能的光伏消纳能力。最后,依据光伏修正功率与普通负荷变化确定了储能蓄电池的容量。结果表明:采用Elman神经网络的修正功率误差可保持在±1.5 k W以内。昼夜温度在10~20℃区间变化时,VRLA蓄电池的相对容量的变化区间为92.6%~97.1%。当地200 k Wp的光伏电站每日约产生485.786 k W·h的光伏发电量可供储能消纳。

To address the problem of missing data caused by the imperfect collection devices at the early stage of the construction of the photovoltaic storage plant, Elman feedforward neural network is used to forecast PV power curve to ensure the reliability and integrity of the data. In addition, the Boltzmann function was used to fit the curve of the relative capacity of the battery with the ambient temperature to consider the impact of ambient temperature changes on the relative capacity of the storage battery and ensure the PV consumption capacity of the storage. Finally, the capacity of the storage battery was determined based on the PV correction power and the general load variation. The results show that the correction power error using Elman neural network can be kept within ±1.5kW. The relative capacity of the VRLA battery varied between 92.6% and 97.1% when the diurnal temperature varied in the range of 10℃ to 20℃. The local200kWp PV plant generates approximately 485.786kW·h of PV power per day that can be consumed by energy storage.