基于多尺度形态学分析的风速预测

Wind speed forecasting based on multi-scale morphological analysis

采用数学形态学和支持向量回归相结合的方法对提前1h间隔为10min的风速预测进行研究。用基于形态学的自适应多尺度算法将原始风速序列分解成一系列具有不同频率和波形特征的细节分量和滤波后主分量,用支持向量回归算法分别对这些分量进行预测,将各预测结果叠加得到最终预测结果。以某风电场的实测风速作为应用案例,实验结果表明,分解后的分量内部规律性更强,与分解前相比,预测精度有显著提高。所提方法为风速预测开辟了一条新的思路,所建MM-SVR模型在实际中有较大推广应用价值。

This paper studies the wind speed prediction for 1 h ahead at a resolution of every 10 min by means of mathematical morphology and support vector regression.At first,the original wind speed sequences are decomposed into a series of subsequences with different frequencies and wave characters by adaptive multi-scale morphological algorithm.Then we predict the subsequences with the method of SVR respectively.At last,the final predicted wind speed can be calculated by the superposition of respective predictions.According to the application case of the actual measured wind speeds in a wind farm,the results indicate that the internal laws of the subsequences are strengthened.Compared with the one which was not decomposed,the prediction accuracy is significantly improved.The proposed method provides a new way for wind speed forecasting,and the MM-SVR model has a big promotion and application value in practice.