基于遗传算法—模糊径向基神经网络的光伏发电功率预测模型

针对光伏发电系统出力波动问题,提出遗传算法(GA)—模糊径向基(RBF)神经网络的光伏发电功率预测模型,将功率预测值应用于光伏发电的蓄电池储能功率调节系统,以降低对电网的冲击。选择与待预测日天气类型相同、日期相近、温度欧氏距离最小的历史日作为相似日,把与光伏发电功率相关性大的太阳辐射强度和温度作为模型输入变量,提出K均值聚类和遗传算法的参数优化方法,建立基于GA—模糊RBF神经网络的最终预测模型。在光伏功率预测的基础上,提出一种平滑控制策略,对光伏并网功率进行有效调节,从而达到平滑光伏功率波动的目的。实例证明,所述预测模型具有较高精度,并验证了平滑功率波动控制策略的有效性。

Photovoltaic Power Forecasting Model Based on Genetic Algorithm and Fuzzy Radial Basis Function Neural Network

To deal with the problem of fluctuating photovoltaic power, a photovoltaic power forecasting model based on genetic algorithm (GA) and fuzzy radial basis function (RBF) neural network is proposed and the output power is applied to the battery energy storage system to mitigate the electric shock on the power system. A historical day of identical weather type, a close date and minimum temperature Euclidean distance is chosen as the similar day. The solar radiation intensity and temperature closely correlated with photovoltaic (PV) power are chosen as input variables of the model, a GA and fuzzy RBF neural network is built as the final prediction model based on the parameter optimization method of K-means clustering and genetic algorithm. Furthermore, a smooth control strategy considering PV power forecasting is used to control the grid-connected PV power, so as to smooth the PV power fluctuation. The experimental results show that the proposed forecasting model has high accuracy and the smooth control strategy for power fluctuation based on photovoltaic power forecasting is effective. This work is supported by Specialized Research Fund for the Doctoral Program of Higher Education (SRFDP) of China (No. 20110008110042) and National Natural Science Foundation of China (No. 51477174).