基于近邻传播聚类与LSTNet的分布式光伏电站群短期功率预测

Short-term Power Forecasting of Distributed Photovoltaic Station Clusters Based on Affinity Propagation Clustering and Long Short-term Time-series Network

为了应对分布式光伏渗透率不断提高带给电网运行的挑战,提出了一种基于近邻传播聚类与长短期时间序列网络(LSTNet)的区域分布式光伏电站群短期功率预测模型。首先,利用近邻传播算法划分区域内不同季节的分布式光伏电站群,并通过皮尔逊相关系数确定光伏出力的强相关气象因子,结合双线性插值法加密对应光伏电站群的气象数据。然后,通过LSTNet挖掘光伏功率和气象因子序列的长期和短期时空依赖,并叠加自回归的线性分量,实现了群内多个光伏电站的同时预测。最后,利用美国国家能源部可再生能源实验室的实测数据集验证了所提方法的有效性。实验比较表明,所提预测模型具有较高的预测精度和鲁棒性。

To cope with the challenges brought by the increasing penetration rate of distributed photovoltaic to power grid operation, a short-term power forecasting model of regional distributed photovoltaic station clusters based on affinity propagation clustering and long short-term time-series network(LSTNet) is proposed. First, the affinity propagation algorithm is employed to divide the regional distributed photovoltaic station clusters into different seasons. Pearson correlation coefficient is used to determine the strong correlation between meteorological factors of photovoltaic output, and the bilinear interpolation method is utilized to encrypt the meteorological data of the corresponding photovoltaic station cluster. Furthermore, LSTNet is used to mine the long-term and short-term temporal and spatial dependence of photovoltaic power and meteorological factor series, and linear components of autoregression are superimposed to realize the simultaneous forecasting of multiple photovoltaic stations in the group. Finally, the effectiveness of the proposed method is verified on the measured data set from the National Renewable Energy Laboratory(NREL) of USA. The experimental comparison shows that the forecasting model obtains high forecasting accuracy and robustness.