Aiming at the problems caused by the current indirect methods for ultra-short-term wind speed prediction which used the same model for each frequency sequence, this paper proposed a combined ultra-short-term wind speed prediction method using a hybrid model based on wavelet decomposition. The proposed method is based on three basic methods, which are autoregressive differential moving average model, back propagation (BP) neural network and support vector machine. The appropriate method is selected and the corresponding model is created according to the characteristics of each frequency sequence after wavelet decomposition, then the results of ultra-short-term wind speed prediction can be obtained through wavelet reconstruction. The proposed method can consider the difference and predictability among each frequency sequence fundamentally, therefore the prediction accuracy can be improved. The proposed method has higher prediction accuracy under different time horizons. When mean absolute error is used as the evaluation index, the prediction accuracy of the proposed method can be improved by 64.2% and 61.4% under 1 h and 4 h time horizons respectively compared with the persistence method, 7.2% and 5.7% under 1 h and 4 h time horizons respectively compared with the minimum prediction error method in traditional single model combination prediction method.
郑若楠. 基于小波分解的超短期风速混合模型组合预测[J]. 分布式能源, 2018, 3(6): 38-46.
ZHENG Ruonan. Combination Prediction Method of Ultra-Short-Term Wind Speed by Using A Hybrid Model Based on Wavelet Decomposition[J]. Distributed Energy, 2018, 3(6): 38-46.
表1 持续法与直接预测法预测结果(1 h预测时长) Table 1 Results of persistence method and direct prediction methods (1 h time horizon)m/s
预测方法
风电场1
风电场2
RMSE
MAE
RMSE
MAE
持续法
1.15
0.80
0.91
0.65
ARIMA
1.14
0.81
0.90
0.65
BPNN
1.14
0.82
0.95
0.69
SVM
1.14
0.81
0.95
0.68
表1 持续法与直接预测法预测结果(1 h预测时长)
图4 在不同分解层数下各种预测方法的预测精度(1 h预测时长)
表2 在最佳小波分解层数下各种风速预测方法的预测精度(1 h预测时长) Table 2 Comparison of wind speed prediction accuracy under the best decomposition level(1h time horizon)m/s
预测方法
风电场1
风电场2
RMSE
MAE
RMSE
MAE
ARIMA
0.63
0.44
0.62
0.43
BPNN
0.44
0.30
0.42
0.26
SVM
0.48
0.32
0.40
0.26
混合模型
0.41
0.28
0.38
0.24
表2 在最佳小波分解层数下各种风速预测方法的预测精度(1 h预测时长)
图5 在最佳小波分解层数下各种风速预测方法的预测精度月变化曲线(1 h预测时长)
表3 持续法与直接预测法预测结果(4 h预测时长) Table 3 Results of persistence method and direct prediction methods (4 h time horizon)(m/s)
预测方法
风电场1
风电场2
RMSE
MAE
RMSE
MAE
持续法
2.03
1.47
1.64
1.16
ARIMA
1.95
1.43
1.60
1.14
BPNN
1.95
1.47
1.62
1.19
SVM
1.97
1.48
1.61
1.17
表3 持续法与直接预测法预测结果(4 h预测时长)
表4 在最佳小波分解层数下各种风速预测方法的预测精度(4 h预测时长) Table 4 Comparison of wind speed prediction accuracy under the best decomposition level(4 h time horizon)m/s
预测方法
风电场1
风电场2
RMSE
MAE
RMSE
MAE
ARIMA
0.98
0.69
1.31
0.93
BPNN
0.81
0.57
0.76
0.50
SVM
0.85
0.59
0.79
0.53
混合模型
0.75
0.54
0.72
0.47
表4 在最佳小波分解层数下各种风速预测方法的预测精度(4 h预测时长)
图6 在不同分解层数下各种预测方法的预测精度(4 h预测时长)
图7 在最佳小波分解层数下各种风速预测方法的预测精度月变化曲线(4 h预测时长)
[1]
国家能源局. 风电发展“十三五”规划[EB/OL]. [2016-11].
[2]
REN Guorui, LIU Jinfu, WAN Jie, et al. Overview of wind power intermittency: impacts, measure-ments, and mitigation solutions[J]. Applied Energy, 2017, 204: 47-65.
[3]
XUE Yusheng, YU Chen, ZHAO Junhua, et al. A review on short-term and ultra-short-term wind power prediction[J]. Automation of Electric Power Systems, 2015, 39(6): 141-151.
JIANG Yuewen, WEN Buying. Real-time market dispatch based on ultra-short-term forecast error of wind power[J]. Electric Power Automation Equipment, 2015, 35(3): 12-17.
KANG Li. Complex terrain wind farm super short term wind resources prediction research[D]. Hohhot: Inner Mongolia University of Technology, 2017.
[5]
康丽. 复杂地形风电场超短期风资源预测研究[D]. 呼和浩特：内蒙古工业大学，2017.
[6]
KAVASSERI R G, SEETHARAMAN K. Day-ahead wind speed forecasting using f-ARIMA models[J]. Renewable Energy, 2009, 34(5): 1388-1393.
[7]
GEORGE G, EVGENIA P, ARISTOTELIS L. A hybrid bayesian kalman filter and applications to numerical wind speed modeling[J]. Journal of Wind Engineering and Industrial Aerodynamics, 2017, 167: 1-22.
[8]
HAN Shuang, MENG Hang, LIU Yongqian, et al. Study on optimization of BP neural network wind power prediction model with two hidden layers[J]. Acta Energiae Solaris Sinica, 2015, 36(9): 2238-2244.
üMMüHAN B, TANSU F. Wind speed prediction using artificial neural networks based on multiple local measurements in Eskisehir[J]. Energy Procedia, 2017, 107: 264-269.
[10]
WANG Han, YAN Jie, LIU Yongqian, et al. Multi-step-ahead method for wind speed prediction correction based on numerical weather prediction and historical measurement data[C]//Wind Europe Conference and Exhibition 2017. Amsterdam, Netherlands. Netherlands Wind Energy Association(NWEA), 2017: 012007.
[11]
HAN Zhonghe, LI Qiuju, YUAN Yiming, et al. RVM based short-term wind speed prediction model[J]. Journal of Electric Power Science and Technology, 2017, 32(3): 38-42.
DING Teng, FENG Donghan, LIN Xiaofan, et al. Ultra-short-term wind speed forecasting based on improved ARIMA-GARCH model[J]. Power System Technology, 2017, 41(6): 1808-1814.
JIANG P, WANG Y, WANG J. Short-term wind speed forecasting using a hybrid model[J]. Energy, 2016, 119: 561-577.
[14]
LIU H, TIAN H Q, PAN D F, et al. Forecasting models for wind speed using wavelet, wavelet packet, time series and Artificial Neural Networks[J]. Applied Energy, 2013, 107(4): 191-208.
[15]
GAO Yang, ZHONG Hongyu, CHEN Xinyu, et al. Ultra short term wind speed forecasting based on neural network and wavelet analysis[J]. Renewable Energy Resources, 2016, 34(5): 705-711.
ZHANG Yan, HAN Pu, WANG Dongfeng, et al. Short-termprediction of wind speed for wind farm based on variational mode decomposition and LSSVM model[J]. Acta Energiae Solaris Sinica, 2018, 39(1): 194-202.
