地球科学进展 ›› 2021, Vol. 36 ›› Issue (9): 937 -949. doi: 10.11867/j.issn.1001-8166.2021.090

研究论文 上一篇    下一篇

基于变分模态分解的机器学习模型择优风速预测系统
摆玉龙( ),路亚妮,刘名得   
  1. 西北师范大学物理与电子工程学院,甘肃 兰州 730070
  • 收稿日期:2021-05-20 修回日期:2021-08-01 出版日期:2021-09-10
  • 基金资助:
    国家自然科学基金项目“基于尺度空间理论和地统计学的数据同化观测误差研究”(41861047)

Wind Speed Forecasting System Based on Variational Mode Decomposition and the Optimal Machine Learning Models

Yulong BAI( ),Yani LU,Mingde LIU   

  1. College of Physics and Electrical Engineering,Northwest Normal University,Lanzhou 730070,China
  • Received:2021-05-20 Revised:2021-08-01 Online:2021-09-10 Published:2021-10-15
  • About author:BAI Yulong (1973-), male, Huining County, Gansu Province, Professor. Research areas include data assimilation and control theory and applications. E-mail: baiyulong@nwnu.edu.cn
  • Supported by:
    the National Natural Science Foundation "Research on observation error of data assimilation based on scale space theory and geostatistics"(41861047)

精准的风速预报对风力发电系统具有重要意义,但风速信号自身固有的随机性使其波动复杂且不可控,以往的研究采用单一或固定的组合模型很难把握风速序列的特征。提出一种基于分解的机器学习模型择优风速预测系统,采用变分模态分解算法降低原始风速序列的复杂度。进而利用模糊神经网络、非线性自回归神经网络、Elman神经网络、反向传播神经网络和自回归差分移动平均模型构成机器学习模型择优系统,分别对子序列的验证集进行预测,通过均方根误差等性能指数选择其最优模型,提高了整体模型的预测精度。试验采用宁夏地区4个站点的实测风速数据,仿真实验结果表明,所提模型相比于单模型以及较新的深度学习组合模型,具有更高的预测精度。

Due to the limited reserves of traditional energy and the excessive use of non-renewable energy, the global carbon emissions seriously accelerate the global greenhouse effect. Compared with traditional energy, wind energy is a kind of green renewable energy. Accurate wind speed prediction is of great significance to wind power generation system, but the inherent randomness of wind speed signal makes its fluctuation complex and uncontrollable. It is difficult to grasp the characteristics of wind speed series using single or fixed hybrid models in previous studies. Therefore, a wind speed forecasting system is proposed based on decomposition and the optimal machine learning models. First, the Variational Mode Decomposition (VMD) algorithm is used to decompose the original wind speed sequence into several sub-sequences. Then, five models of Fuzzy Neural Network (FNN), Nonlinear Auto-regressive Neural Network (NARX), Elman neural network, Back Propagation Neural Network (BPNN) and Auto-regressive integrated Moving Average model (ARIMA), are selected to construct a machine learning model selection system and predict the validation set of sub-sequences, respectively. The root mean square error (RMSE) is used to select the optimal model and the final prediction result is obtained by summing up the predicted values of each sub-sequence. The measured wind speed data from four stations in Ningxia Hui Autonomous Region was used in the experiment. The sampling height was 70 meters and the time interval was 15 minutes. Three groups of comparison models were established to verify the prediction effect of the proposed model. The first group: five single models used in this experiment: FNN, NARX, Elman, BPNN and ARIMA; The second group: five hybrid models based on VMD decomposition: VMD-FNN, VMD-NARX, VMD-Elman, VMD-BPNN and VMD-ARIMA; The third group: newer hybrid models of deep learning: Improved Complete Ensemble Empirical Mode Decomposition (ICEEMDAN)-Long Short-Term Memory (LSTM) and ICEEMDAN-Gated Recurrent Unit (GRU). The experimental results indicate that the proposed model is superior to other baseline models. The VMD decomposition method can effectively reduce the complexity of the original wind speed sequence. For different sub-sequences after decomposition, the optimal method of machine learning model can capture the sequence features and obtain the optimal prediction results.

中图分类号: 

图1 FNN模型的系统框图
Fig. 1 System chart of FNN
图2 Elman神经网络结构图
Fig. 2 Structure chart of Elman neural network
图3 基于VMD择优系统的预测框图
Fig. 3 Flowchart of a prediction system based on VMD and machine learning
图4 数据位置信息图及风速序列曲线
Fig. 4 Data position information graph and wind speed sequence curve
表1 风速数据特征信息
Table 1 Wind speed data characteristic information
图5 序列1VMD分解得到的子序列
Fig. 5 Sub-sequence 1 obtained by VMD decomposition
表2 序列 1各子序列验证集模型预测的 RMSE指标
Table 2 Each sub-sequence verifies the RMSE index predicted by the model set of sequence 1
图6 各子序列测试集的预测结果
Fig. 6 The prediction results of each sub-sequence
图7 4组风速数据预测对比
Fig. 7 Comparison of four groups of wind speed data forecast
图8 4组序列的误差指标图:MAPERMSEMAE
Fig. 8 The error indicators of four groups of sequences were MAPE RMSE and MAE
表3 各序列相关预测误差指标
Table 3 Correlation prediction error index of each sequence
序列 模型 RMSE/(m/s) MAPE/% MAE/(m/s) 序列 模型 RMSE/(m/s) MAPE/% MAE/(m/s)

1

FNN 0.6463 18.35 0.5094

2

FNN 0.6896 8.13 0.4686
NARX 0.6563 19.33 0.5212 NARX 0.5950 8.31 0.4519
Elman 0.6127 16.15 0.4871 Elman 0.5625 7.85 0.4210
BPNN 0.6213 17.32 0.4947 BPNN 0.5625 7.89 0.4192
ARIMA 0.5491 13.66 0.3019 ARIMA 0.4803 7.22 0.3734
VMD-FNN 0.7016 27.11 0.4971 VMD-FNN 0.4711 6.82 0.3623
VMD-NARX 0.3156 9.74 0.2453 VMD-NARX 0.2737 4.06 0.2127
VMD-Elman 0.3511 10.58 0.2684 VMD-Elman 0.2975 4.46 0.2344
VMD-BPNN 0.5457 13.63 0.4348 VMD-BPNN 0.5633 8.53 0.4301
VMD-ARIMA 0.2996 7.02 0.2337 VMD-ARIMA 0.2688 3.90 0.2099
ICEEMDAN-LSTM [ 36 ] 0.3306 7.85 0.2619 ICEEMDAN-LSTM [ 36 ] 0.3990 5.96 0.3153
ICEEMDAN-GRU[ 36 ] 0.2412 5.58 0.1914 ICEEMDAN-GRU [ 36 ] 0.2246 3.52 0.1727
VMD-NARX-ARIMA 0.2982 6.83 0.2319 VMD-NARX-ARIMA 0.2670 3.87 0.2091

3

FNN 1.2672 20.67 0.6966

4

FNN 1.7746 31.73 1.3013
NARX 0.5936 12.82 0.1666 NARX 1.6968 29.98 1.2249
Elman 0.6365 11.72 0.3436 Elman 1.6806 28.50 1.2248
BPNN 0.4428 18.47 0.1894 BPNN 1.5349 25.85 1.1403
ARIMA 0.4176 10.01 0.1582 ARIMA 1.3684 21.69 0.9719
VMD-FNN 1.6047 23.25 1.0088 VMD-FNN 1.2072 18.59 0.8476
VMD-NARX 1.1591 11.56 0.6276 VMD-NARX 0.3856 5.03 0.2208
VMD-Elman 1.2498 14.03 0.7190 VMD-Elman 0.4460 5.96 0.3355
VMD-BPNN 0.4059 15.35 0.2300 VMD-BPNN 1.4769 24.65 1.1013
VMD-ARIMA 0.1948 5.14 0.0864 VMD-ARIMA 0.3425 4.55 0.2643
ICEEMDAN-LSTM [ 36 ] 0.2952 8.62 0.2131 ICEEMDAN-LSTM [ 36 ] 1.0671 16.24 0.8088
ICEEMDAN-GRU[ 36 ] 0.2223 5.76 0.0758 ICEEMDAN-GRU [ 36 ] 0.7932 11.96 0.6007
VMD-NARX-ARIMA 0.1936 5.03 0.0851 VMD-NARX-ARIMA 0.3374 4.41 0.2627
1 IRENA. Wind energy [EB/OL]. 2021. .
URL    
2 LIU H, CHEN C. Data processing strategies in wind energy forecasting models and applications: a comprehensive review[J]. Applied Energy, 2019, 249(SEP.):392-408.
3 JUNG J, BROADWATER R P. Current status and future advances for wind speed and power forecasting[J]. Renewable & Sustainable Energy Reviews, 2014, 31(MAR.):762-777.
4 MA Leiming. Development of artificial intelligence technology in weather forecast[J]. Advances in Earth Science, 2020, 35(6):551-560.
马雷鸣. 天气预报中的人工智能技术进展[J]. 地球科学进展,2020,35(6): 551-560.
5 CAO B, ZHAO J W, Lü Z H, et al. Multi-objective evolution of fuzzy rough neural network via distributed parallelism for stock prediction[J]. IEEE Transactions on Fuzzy Systems, 2020, 28(5): 939-952.
6 WANG Bingdi, LI Qingquan, SHEN Xinyong, et al. Climatological characteristics of the East Asian winter monsoon simulated by CWRF regional climate model[J]. Advances in Earth Science, 2020, 35(3):319-330.
王冰笛,李清泉,沈新勇,等. 区域气候模式 CWRF 对东亚冬季风气候特征的模拟[J]. 地球科学进展,2020,35(3):319-330.
7 HOOLOHAN V, TOMLIN A S, COCKERILL T. Improved near surface wind speed predictions using Gaussian process regression combined with numerical weather predictions and observed meteorological data[J]. Renewable Energy, 2018, 126 (OCT.) :1 043-1 054.
8 ALLEN D J, TOMLIN A S, BALE C S E, et al. A boundary layer scaling technique for estimating near-surface wind energy using numerical weather prediction and wind map data[J]. Applied Energy, 2017, 208(15):1 246-1 257.
9 Guoqin Lü, ZHANG Huiqin, LI Jiachun. Characteristics and forecast of short-term wind speed series in the Donghai Bridge wind farm [J]. Scientia Sinica (Physica, Mechanica & Astronomica), 2016, 46(12):124713.
吕国钦,张会琴,李家春. 东海大桥风电场短期风速序列特性及其预报[J]. 中国科学:物理学力学天文学,2016,46(12):124713.
10 LIU H, TIAN H Q, LI Y F. Comparison of two new ARIMA-ANN and ARIMA-Kalman hybrid methods for wind speed prediction[J]. Applied Energy, 2012, 98(OCT.): 415-424.
11 WANG L, LI X, BAI Y. Short-term wind speed prediction using an extreme learning machine model with error correction[J]. Energy Conversion and Management, 2018, 162(APR.):239-250.
12 SOMAN S, ZAREIPOUR H, MALIK O, et al. A review of wind power and wind speed forecasting methods with different time horizons[C]. North American Power Symposium (NAPS), USA: IEEE, 2010.
13 NOOROLLAHI Y, JOKAR M A, KALHOR A. Using artificial neural networks for temporal and spatial wind speed forecasting in Iran[J]. Energy Conversion and Management, 2016, 115(May.): 17-25.
14 KONAR A, BHATTACHARYA D. Time-series prediction and application: a machine intelligence approach[M]. Intelligent System Reference Library,2017
15 PARK C, LEE C, HONG L, et al. S2-Net: machine reading comprehension with SRU-based self-matching networks[J]. ETRI Journal, 2019, 41(3): 371-382.
16 BENMOUIZA K, CHEKNANE A. Small-scale solar radiation forecasting using ARMA and nonlinear autoregressive neural network models[J]. Theoretical & Applied Climatology, 2016, 124(3/4): 945-958.
17 YU C, LI Y, XIANG H, et al. Data mining-assisted short-term wind speed forecasting by wavelet packet decomposition and elman neural network[J]. Journal of Wind Engineering and Industrial Aerodynamics, 2018, 175(APR.): 136-143.
18 LIU M D, DING L, BAI Y L. Application of hybrid model based on empirical mode decomposition, novel recurrent neural networks and the ARIMA to wind speed prediction[J]. Energy Conversion and Management, 2021, 233(FEB.): 113917.
19 TANG G, WU Y, LI C, et al. A novel wind speed interval prediction based on error prediction method[J]. IEEE Transactions on Industrial Informatics, 2020, 16(11): 6 806-6 815.
20 WANG Jun, LI Xia, ZHOU Xidong, et al. Ultra-short-term wind speed prediction based on VMD-LSTM[J]. Power System Protection and Control, 2020, 48(11): 45-52.
王俊,李霞,周昔东,等. 基于VMD和LSTM的超短期风速预测[J]. 电力系统保护与控制,2020,48(11):45-52.
21 LIU H, MI X, LI Y. Smart multi-step deep learning model for wind speed forecasting based on variational mode decomposition, singular spectrum analysis, LSTM network and ELM [J]. Energy Conversion and Management, 2018, 159(MAR.): 54-64.
22 FEI S W, HE Y. Wind speed prediction using the hybrid model of wavelet decomposition and artificial bee colony algorithm-based relevance vector machine [J]. International Journal of Electrical Power & Energy Systems, 2015, 73 (DEC.): 625-631.
23 SANTHOSH M, VENKAIAH C, KUMAR D. Ensemble empirical mode decomposition based adaptive wavelet neural network method for wind speed prediction [J]. Energy Conversion and Management, 2018, 168(15): 482-493.
24 TIAN Z. Short-term wind speed prediction based on LMD and improved FA optimized combined kernel function LSSVM[J]. Engineering Applications of Artificial Intelligence, 2020, 91(MAY.): 103573.
25 CHEN Zhuan, XIONG Xin, YOU Yuyin. Variational mode decomposition and long and short time neural network for dam deformation prediction[J]. Science of Surveying and Mapping, 2021, 49(9):34-42.
陈竹安,熊鑫,游宇垠. 变分模态分解与长短时神经网络的大坝变形预测[J]. 测绘科学,2021, 49(9):34-42.
26 LIU Yu, ZHANG Yufei, CHEN Shangqiao. Ultra-short-term wind speed prediction based on variational mode decomposition and bidirectional long short-term memory neural network[J]. Industrial Control Computer, 2020, 33(9): 54-57.
刘宇,张雨飞,陈尚巧. 基于变分模态分解与双向长短期记忆神经网络的超短期风速预测[J]. 工业控制计算机,2020,33(9):54-57.
27 ZHANG Y, PAN G, CHEN B, et al. Short-term wind speed prediction model based on GA-ANN improved by VMD[J]. Renewable Energy, 2020, 156(AUG.): 1 373-1 388.
28 Hu H L, Wang L, Tao R. Wind speed forecasting based on variational mode decomposition and improved echo state network[J]. Renewable Energy, 2021, 164(FEB.): 729-751.
29 LIU H, YANG R, WANG T, et al. A hybrid neural network model for short-term wind speed forecasting based on decomposition, multi-learner ensemble, and adaptive multiple error corrections[J]. Renewable Energy, 2021, 165(1): 573-594.
30 ZHANG Y G, YUAN Z, KONG C H, et al. A new prediction method based on VMD-PRBF-ARMA-E model considering wind speed characteristic[J]. Energy Conversion and Management, 2020, 203(1): 112254.
31 DRAGOMIRETSKIY K, ZOSSO D. Variational mode decomposition[J]. IEEE Transactions on Signal Processing, 2014, 62(3): 531-544.
32 YU Sheng, ZHOU Hongbo, YU Fan, et al. Application of Fuzzy Neural Network in power short-term load forecasting[J]. Power Grid Analysis & Study, 2018, 46(11): 88-91.
喻圣,邹红波,余凡,等. 模糊神经网络在电力系统短期负荷预测中的应用[J]. 电网分析与研究,2018,46(11):88-91.
33 TANG L H, BAI Y L, YANG J, et al. A hybrid prediction method based on empirical mode decomposition and multiple model fusion for chaotic time series[J]. Chaos, Solitons & Fractals, 2020, 141(DEC.): 110366.
34 YI Xuejun, WEI Shouke, SHI Yuhao, et al. Hydrological series prediction model based on wavelet nonlinear autoregressive network[J]. Computer Technology and Development, 2021, 31(3): 70-77.
衣学军,魏守科,石玉好,等. 基于小波非线性自回归网络的水文预测模型[J]. 计算机技术与发展,2021,31(3):70-77.
35 GUO Mingxin, HUANG Ruanming, BIAN Xiaoyan, et al. Shortterm wind speed time series forecasting method based on elman neural network and software development[J]. Industrial Control Computer, 2021, 34(2): 83-85.
郭明星,黄阮明,边晓燕,等. 基于Elman神经网络的短期风速时间序列预测及软件开发[J]. 工业控制计算机,2021,34(2):83-85.
36 DUAN J K, ZUO H C, BAI Y L, et al. Short-term wind speed forecasting using recurrent neural networks with error correction[J]. Energy, 2021, 217(15): 119397.
37 HE Y, LI J M, SUMEI R, et al. A hybrid model for financial time series forecasting-integration of EWT, ARIMA with the improved ABC optimized ELM[J]. IEEE Access, 2020, 8: 84 501-84 518.
[1] 张克存, 屈建军, 鱼燕萍, 韩庆杰, 王涛, 安志山, 胡菲. 中国铁路风沙防治的研究进展[J]. 地球科学进展, 2019, 34(6): 573-583.
[2] 祁建华, 李孟哲, 高冬梅, 甄毓, 张大海. 沙尘天气对大气生物气溶胶中微生物浓度、特性和分布的影响[J]. 地球科学进展, 2018, 33(6): 568-577.
[3] 陈晓龙, 吴波, 周天军. FGOALS-s2海洋同化系统中东亚夏季风和前冬厄尔尼诺—南方涛动关系的年代际变化[J]. 地球科学进展, 2017, 32(4): 362-372.
[4] 孙倩, 吴波, 周天军. 基于可预测模态分析技术的亚澳夏季风统计—动力季节预测模型及其回报技巧评估[J]. 地球科学进展, 2017, 32(4): 420-434.
[5] 郝青振, 张人禾, 汪品先, 王斌. 全球季风的多尺度演化[J]. 地球科学进展, 2016, 31(7): 689-699.
[6] 陆雯茜, 吴涧. 气溶胶影响印度夏季风和东亚夏季风的研究进展[J]. 地球科学进展, 2016, 31(3): 248-257.
[7] 王训明, 周娜, 郎丽丽, 花婷, 焦琳琳, 马文勇. 风沙活动对陆地生态系统影响研究进展[J]. 地球科学进展, 2015, 30(6): 627-635.
[8] 邵勰, 黄平, 黄荣辉. 南海夏季风爆发的研究进展[J]. 地球科学进展, 2014, 29(10): 1126-1137.
[9] 张琳琳, 赵晓英, 原慧. 风对植物的作用及植物适应对策研究进展[J]. 地球科学进展, 2013, 28(12): 1349-1353.
[10] 李宁,顾孝天,刘雪琴. 沙尘暴灾害致灾因子三维联合分布与重现期探索[J]. 地球科学进展, 2013, 28(4): 490-496.
[11] 杨秋明,宋娟,李熠,谢志清,黄世成,钱玮. 全球大气季节内振荡对长江流域持续暴雨影响的研究进展[J]. 地球科学进展, 2012, 27(8): 876-884.
[12] 林祥,钱维宏. 全球季风和季风边缘研究[J]. 地球科学进展, 2012, 27(1): 26-34.
[13] 胡金明,何大明,李运刚. 从湿季降水分异论哀牢山季风交汇[J]. 地球科学进展, 2011, 26(2): 183-192.
[14] 张克存,屈建军,牛清河,张伟民,韩庆杰. 青藏铁路沿线砾石方格固沙机理风洞模拟研究[J]. 地球科学进展, 2010, 25(3): 284-289.
[15] 邓振镛,张强,徐金芳,陈敏,秦三杰,张树誉. 高温热浪与干热风的危害特征比较研究[J]. 地球科学进展, 2009, 24(8): 865-873.
阅读次数
全文


摘要