基于风速时空关联的多步预测方法

MULTI-STEP WIND SPEED PREDICTION METHOD BASED ON WIND SPEED SPATIAL-TIME CORRELATION

基于风速的空间关联性提出一种新的多位置多步风速组合预测方法。对风场内各风力机进行灰色关联分析,并据此利用昆虫优化算法进行优选重构,获取目标风力机及临近域空间信息。利用卷积神经网络对重构矩阵进行空间特征提取,并输入长短时记忆网络进行多步预测。最后,将所提方法应用于不同风场进行风速预测,通过对比分析验证所提方法的预测精度和泛化能力。

Ultra-short-term prediction for considering the spatial correlation of wind speed is a research hotspot for large-scale wind power grid connection. This paper proposes a new multi-position and multi-step combined wind speed prediction method based on the spatial correlation of wind speed. The grey correlation analysis on each wind turbine in the wind farm is carried out,and then the pity beetle algorithm is adopted to perform optimal and reconstruct to obtain the target wind turbine and its adjacent space information. The convolutional neural network is used to extract the spatial features of the reconstructed matrix,which are input the long-short-term memory networks for multi-step prediction. Finally,the method in this paper is applied to different wind farms for wind speed prediction.The prediction accuracy and generalization ability of the proposed method are verified through comparative analysis.