基于分形优化的VMD和GA-BP的短期风速预测  

SHORT-TERM WIND SPEED PREDICTION BASED ON FRACTAL OPTIMIZATION OF VMD-GA-BP

作　　者：全一鸣 喻敏[1,2] 王文波 魏来[1,2] QuanYiming;Yu Min;Wang Wenbo;Wei Lai(College of Science,Wuhan University of Science and Technology,Wuhan 430065,China;Hubei Province Key Laboratory of Systems Science in Metallurgical Process,Wuhan University of Science and Technology,Wuhan 430065,China)

机构地区：[1]武汉科技大学理学院,武汉430065 [2]冶金工业过程系统科学湖北省重点实验室(武汉科技大学),武汉430065

出　　处：《太阳能学报》2023年第7期436-446,共11页Acta Energiae Solaris Sinica

基　　金：国家自然科学基金(61671338);冶金工业过程系统科学湖北省重点实验室基金重点项目(Y202007,Z201901);大学生创新训练项目(S202110488050)。

摘　　要：该文首次提出基于分形优化的变分模态分解(VMD)和遗传算法(GA)改进的反向传播神经网络(BP)模型的短期风速预测方法。首先使用计盒维数算法优化VMD分解层数,然后针对风速序列的非平稳性,利用优化后的VMD分解原始风速序列得到较平稳风速子序列,最后采用遗传算法改进的BP神经网络分别训练预测各模态分量,并通过叠加所有分量预测值得到最终预测结果。使用该方法对美国某风电场风速进行预测,将预测结果与BP、VMD-ARMA、VMD-LSTM、VMD-BP、基于分形优化VMD-BP模型对比,并选取MAE、RMSE、MAPE这3种评价指标分别评价上述6个模型。结果表明:使用基于分形优化的VMD-GA-BP模型能显著提高预测效果,降低风速预测误差。In the background of the sharp reduction of traditional energy sources,there is an urgent need to propose an accurate wind speed prediction method to ensure the normal operation of power systems.This paper proposes for the first time the variational mode decomposition(VMD)and genetic algorithm(GA)based on fractal optimization.GA improved back propagation(BP)neural network model for short-term wind speed prediction.Firstly,the box-counting dimension algorithm was used to optimize the decomposition layers of VMD.Then,aiming at the non-stationarity of wind speed sequence,the original wind speed sequence was decomposed by the optimized VMD to obtain a relatively stable wind speed sub-sequence.Finally,the BP neural network improved by genetic algorithm was used to train and predict each modal component respectively,and the final prediction result was obtained by superimposing the predicted values of all components.The wind speed of a wind farm was predicted by this method,and the prediction results were compared with BP,VMD-ARMA,VMD-LSTM,VMD-BP and VMD-BP models based on fractal optimization.MAE,RMSE and MAPE were selected to evaluate the six models.The results show that the VMD-GA-BP model based on fractal optimization can significantly improve the prediction effect and reduce the wind speed prediction error.

关 键 词：分形维数 变分模态分解 反向传播网络 短期风速预测 遗传算法 

分 类 号：O29[理学—应用数学]