摘要
根据青海某5 MW光伏电场的历史光伏发电功率数据和当地的气象预报信息,分析影响功率预测的主要气象因素。采用Elman神经网络算法,结合与预测日同日类型下整点时刻的气象数据和光伏输出功率数据,建立光伏发电短期功率预测模型。对不同日类型的光伏出力的预测结果表明,该短期预测模型具有较高的精度,有助于电网能量的调度,对电力系统的安全稳定运行有积极作用。通过与BP神经网络和非线性状态估计(NSET)算法对比研究表明,Elman神经网络具有更高的预测精度。
According to historical power data and weather forecast information of 5 MW PV farm in Qinghai, main weather factors affecting PV generation power forecast were analyzed. Combining the similar type of weather data and PV output power data as the forecast day in the whole hour, Elman neural network algorithm was used to set up short term PV power prediction model. The forecast results for different day show that the short term prediction model has higher precision, redounds to attemperment of grid electricity, and plays active role for safe and stable operation of electric power system. Compared with BP neural network and nonlinear state estimation algorithm, Elman neural network algorithm has higher accuracy.
出处
《太阳能学报》
EI
CAS
CSCD
北大核心
2017年第6期1560-1566,共7页
Acta Energiae Solaris Sinica
