配置储能装置的光伏预测配网优化研究

Optimal operation of photovoltaic forecast distribution network based on information fusion theory

在分析区域光伏并入配网时交流母线PCC点电压波动机理的前提下,提出一种配置储能装置的光伏预测配网优化运行方法。该方法分析光伏发电输出功率与多种气象因素的相关性,将多种气象因素作为多个信息源处理,运用信息融合理论将其加权为一个综合影响因子λ,建立以λ为输入的BP神经网络预测模型,将模型输出的预测值实时传送给储能装置,采用基于滤波原理的光伏发电输出功率平滑控制,实现光伏功率平滑输出,这种方法可以和分时电价有效结合,既提高配网运行经济性的同时稳定节点电压,降低节点电压越限的可能。研究结果表明:所述预测模型具有较高的预测精度,对配网的安全优化运行有一定作用。

An optimized operation method of PV forecasting distribution network with energy storage device was proposed based on the analysis of the voltage fluctuation mechanism of AC busbar PCC point when regional PV was incorporated into the distribution network.The method analyzed the correlation between photovoltaic power output and various meteorological factors.Treating various meteorological factors as multiple information sources,and using information fusion theory to weight it into a comprehensive impact factorλ,BP neural network prediction was established takingλas an input.The model transmitted the predicted value of the model output to the energy storage device in real time,and adopted the smoothing control of the photovoltaic power generation output power based on the filtering principle to realize the smooth output of the photovoltaic power.The results show that the method can be effectively combined with the time-sharing electricity price,and the distribution network operation is improved.The economical stability of the PCC point voltage at the same time reduces the possibility that the node voltage exceeds the limit.The prediction model has higher prediction accuracy and has a certain effect on the safety optimization operation of the distribution network.