基于样本熵双分解和SSA-LSTM的超短期风速预测  

ULTRA-SHORT-TERM WIND SPEED PREDICTION BASED ON SAMPLE ENTROPY-BASED DUAL DECOMPOSITION AND SSA-LSTM

作　　者：王希 孙娜 苏浩 张楠 张帅 纪捷 Wang Xi;Sun Na;Su Hao;Zhang Nan;Zhang Shuai;Ji Jie(Faulty of Automation,Huaiyin Institute of Technology,Huaian 223003,China;Key Laboratory of Thermo-Fluid Science&Engineering of MOE,Xi’an Jiaotong University,Xi’an 710049,China)

机构地区：[1]淮阴工学院自动化学院,淮安223003 [2]西安交通大学能源与动力工程学院热流科学与工程教育部重点实验室,西安710049

出　　处：《太阳能学报》2025年第3期611-618,共8页Acta Energiae Solaris Sinica

基　　金：江苏省高校自然科学基金(20KJD480003);江苏省双创计划(JSSCBS(2020)31035);江苏省自然科学基金(BK20201069)。

摘　　要：为进一步提高超短期风速预测的准确性,提出一种基于样本熵的双分解和麻雀搜索算法(SSA)改进长短时记忆神经网络(ILSTM)的耦合模型(简记为DILSTM)。首先,利用变分模态分解(VMD)分解原始序列并通过样本熵量化各子序列的复杂性;其次,利用完全自适应噪声集合经验模态分解(CEEMDAN)分解复杂度最高的子序列进一步提取特征过滤噪声;最后,将双分解得到的子序列分别建立DILSTM预测模型,并对所有子序列的预测结果叠加得到最终风速预测结果。真实风场数据实验结果表明,所提模型与LSTM、随机森林(RF)、卷积神经网络(CNN)单一模型相比,R2提升约25%,RMSE降低约65%;同时,与现有同类研究对比,也证实了所提DILSTM模型的优势,该文为提高超短期风速预测精准度提供了一种新的方法。To improve the accuracy of ultra-short-term wind speed prediction,a coupled model named DILSTM is proposed,which combines sample entropy,dual decomposition,and sparrow search algorithm optimized long short-term memory neural network(LSTM).Firstly,the original sequence is decomposed using variational mode decomposition(VMD)and the complexity of each sub sequence is quantified through sample entropy;the subsequence with the highest complexity is further decomposed using complete ensemble empirical mode decomposition with adaptive noise(CEEMDAN),further extracting features and filtering noise;Finally,the sub sequences obtained from the double decomposition are used to establish DILSTM prediction models,and the final wind speed prediction results are obtained by overlaying the prediction results of all sub sequences.The experimental results of real wind field data show the proposed model outperforms single models such as LSTM,random forest(RF),and convolutional neural network(CNN),with an approximately 25%increase in R2 and a 65%decrease in RMSE;Meanwhile,compared with existing similar studies,the advantages of the DILSTM model proposed in this paper have been confirmed.This study provides a new method for improving the accuracy of ultra-short-term wind speed prediction.

关 键 词：风速 神经网络 预测 变分模态分解 样本熵 麻雀搜索算法 

分 类 号：TM614[电气工程—电力系统及自动化]