Ultra-short-term Wind Power Prediction Based on Improved MFO Optimized Attention-LSTM

doi:10.11985/2023.03.038

Abstract

Abstract:

For the uncertainty of wind power, an ultra short-term wind power prediction method based on improved MFO(Moth-flame optimization) attention LSTM(Long short-term memory) is proposed. Firstly, the original power data is decomposed into multiple IMF(Intrinsic mode functions) components by using the adaptive noise complete ensemble empirical mode decomposition (CEEMDAN) to eliminate the influence between different components, calculate the sample entropy of each component, and combine the approximate values of sample entropy to reduce the operation scale. Secondly, by introducing Chebyshev chaotic mapping, Cauchy mutation, Sigmoid function-based inertia weights to improve the traditional moth fire fighting algorithm, and the improved moth fire fighting algorithm is compared with the traditional MFO, PSO(Particle swarm optimization) algorithm. It is proved that its optimization ability has been greatly improved. Finally, the Attention mechanism is used to calculate the different weights of the hidden layer state of the LSTM neural network, and the improved moth-fighting optimization algorithm is used to optimize the hyperparameters of the Attention-LSTM. The combined IMF components are modeled respectively, and each component model is superimposed to obtain the final power prediction curve. Through the simulation analysis of the measured data of a wind farm in Jinzhou, the results show that the proposed model has high prediction accuracy and has certain reference significance for actual engineering.

Key words: Wind power, Attention mechanism, chaos mapping, Cauchy mutation, improved MFO algorithm, long short-term memory

CLC Number:

TM614

SONG Liye, JU Yadong, ZHANG Xin. Ultra-short-term Wind Power Prediction Based on Improved MFO Optimized Attention-LSTM[J]. Journal of Electrical Engineering, 2023, 18(3): 358-368.

Figures/Tables 22

测试函数	维度	搜索范围
$f 1 x = ∑ i = 1 n x i + ∏ i = 1 n x i$	30	[-10，10]
$f 2 x = ∑ i = 1 n x i 4 + random 0, 1$	30	[-1.28，1.28]
$f 3 x = 1 4 000 ∑ i = 1 n x i 2 − ∏ i = 1 n cos x i i + 1$	30	[-600，600]
$f 4 x = ∑ i = 1 n x i − 10 cos 2 π x i + 10$	30	[-5.12，5.12]

References 22

[1]	林卫星, 文劲宇, 艾小猛, 等. 风电功率波动特性的概率分布研究[J]. 中国电机工程学报, 2012, 32(1):38-46.
	LIN Weixing, WEN Jinyu, AI Xiaomeng, et al. Research on probability distribution of wind power fluctuation characteristics[J]. Proceedings of the CSEE, 2012, 32(1):38-46.
[2]	姜贵敏, 陈志军, 李笑竹, 等. 基于EEMD-ACS-LSSVM的短期风电功率预测[J]. 太阳能学报, 2020, 41(5):77-84.
	JIANG Guimin, CHEN Zhijun, LI Xiaozhu, et al. Short-term wind power forecast based on EEMD-ACS-LSSVM[J]. Acta Energiae Solaris Sinica, 2020, 41(5):77-84.
[3]	孙改平, 蒋传文. 基于两级级联聚类的神经网络风电功率预测[J]. 太阳能学报, 2021, 42(3):56-62.
	SUN Gaiping, JIANG Chuanwen. Neural network wind power prediction based on two-level cascade clustering[J]. Acta Energiae Solaris Sinica, 2021, 42(3):56-62.
[4]	黎静华, 黄玉金, 黄乾. 基于改进混沌时间序列的风电功率区间预测方法[J]. 电力自动化设备, 2019, 39(5):53-60,68.
	LI Jinghua, HUANG Yujin, HUANG Qian. Wind power interval forecasting method based onimproved chaotic time series[J]. Electric Power Automation Equipment, 2019, 39(5):53-60,68.
[5]	胡帅, 向月, 沈晓东, 等. 计及气象因素和风速空间相关性的风电功率预测模型[J]. 电力系统自动化, 2021, 45(7):28-36.
	HU Shuai, XIANG Yue, SHEN Xiaodong, et al. Wind power prediction model considering meteorological factor and spatial correlation of wind speed[J]. Automation of Electric Power Systems, 2021, 45(7):28-36.
[6]	杨锡运, 邢国通, 马雪, 等. 一种核极限学习机分位数回归模型及风电功率区间预测[J]. 太阳能学报, 2020, 41(11):300-306.
	YANG Xiyun, XING Guotong, MA Xue, et al. A quantile regression model of nuclear extreme learning machine and interval prediction of wind power[J]. Acta Energiae Solaris Sinica, 2020, 41(11):300-306.
[7]	向玲, 邓泽奇. 基于改进经验小波变换和最小二乘支持向量机的短期风速预测[J]. 太阳能学报, 2021, 42(2):97-103.
	XIANG Ling, DENG Zeqi. Short-term wind speed prediction based on improved empirical wavelet transform and least square support vector machine[J]. Acta Energiae Solaris Sinica, 2021, 42(2):97-103.
[8]	HE Yaoyao, WANG Yun. Short-term wind power prediction based on EEMD-LASSO-QRNN model[J]. Applied Soft Computing Journal, 2021, 105:107288. doi: 10.1016/j.asoc.2021.107288
[9]	谢国民, 倪乐水, 曹媛. 基于VSRP与β-GWO-SVM的变压器故障辨识方法[J]. 高电压技术, 2021, 47(10):3635-3641.
	XIE Guomin, NI Leshui, CAO Yuan. Transformer fault identification method based on VSRP and β-GWO-SVM[J]. High Voltage Engineering, 2021, 47(10):3635-3641.
[10]	LI Yanan, YANG Peng, WANG Huajun. Short-term wind speed forecasting based on improved ant colony algorithm for LSSVM[J]. Cluster Computing, 2019, 22(5):11575-11581. doi: 10.1007/s10586-017-1422-2
[11]	王炜, 刘宏伟, 陈永杰, 等. 基于LSTM循环神经网络的风力发电预测[J]. 可再生能源, 2020, 38(9):1187-1191.
	WANG Wei, LIU Hongwei, CHEN Yongjie, et al. Wind power generation forecast based on LSTM recurrent neural network[J]. Renewable Energy, 2020, 38(9):1187-1191.
[12]	钱勇生, 邵洁, 季欣欣, 等. 基于LSTM-Attention网络的短期风电功率预测[J]. 电机与控制应用, 2019, 46(9):95-100.
	QIAN Yongsheng, SHAO Jie, JI Xinxin, et al. Short-term wind power prediction based on LSTM-Attention network[J]. Electric Machines and Control Application, 2019, 46(9):95-100.
[13]	赵倩, 黄景涛. 基于EMD-SA-SVR的超短期风电功率预测研究[J]. 电力系统保护与控制, 2020, 48(4):89-96.
	ZHAO Qian, HUANG Jingtao. Research on ultra-short-term wind power prediction based on EMD-SA-SVR[J]. Power System Protection and Control, 2020, 48(4):89-96.
[14]	文孝强, 许洋. 基于EMD分解的风力机功率特性分析与预测建模[J]. 太阳能学报, 2021, 42(11):293-298.
	WEN Xiaoqiang, XU Yang. Analysis and prediction modeling of wind turbine power characteristics based on EMD decomposition[J]. Acta Energiae Solaris Sinica, 2021, 42(11):293-298.
[15]	程启明, 陈路, 程尹曼, 等. 基于EEMD和LS-SVM模型的风电功率短期预测方法[J]. 电力自动化设备, 2018, 38(5):27-35.
	CHENG Qiming, CHEN Lu, CHENG Yinman, et al. Short-term wind power forecasting method based on EEMD and LS-SVM model[J]. Electric Power Automation Equipment, 2018, 38(5):27-35.
[16]	YEH J R, SHIEH J S, HUANG N E. Complementary ensemble empirical mode decomposition:A novel noise enhanced data analysis method[J]. Advances in Adaptive Data Analysis, 2010, 2(2):135-156. doi: 10.1142/S1793536910000422
[17]	杨森, 刘三明, 王致杰. 基于GRU-XGBoost的风电场功率短期预测[J]. 仪表技术, 2020(1):17-21.
	YANG Sen, LIU Sanming, WANG Zhijie. Short-term prediction of wind farm power based on GRU-XGBoost[J]. Instrumentation Technology, 2020(1):17-21.
[18]	张学清, 梁军, 张熙, 等. 基于样本熵和极端学习机的超短期风电功率组合预测模型[J]. 中国电机工程学报, 2013, 33(25):33-40.
	ZHANG Xueqing, LIANG Jun, ZHANG Xi, et al. Ultra-short-term wind power combined forecasting model based on sample entropy and extreme learning machine[J]. Proceedings of the CSEE, 2013, 33(25):33-40.
[19]	魏乐, 李思莹. 基于MEEMD-LSSVM的风电功率超短期预测研究[J]. 智慧电力, 2020, 48(5):21-26.
	WEI Le, LI Siying. Research on ultra-short-term prediction of wind power based on MEEMD-LSSVM[J]. Smart Power, 2020, 48(5):21-26.
[20]	潘晓杰, 张立伟, 张文朝, 等. 基于飞蛾扑火优化算法的多运行方式电力系统稳定器参数协调优化方法[J]. 电网技术, 2020, 44(8):3038-3046.
	PAN Xiaojie, ZHANG Liwei, ZHANG Wenchao, et al. Coordinated optimization method of multi-operation power system stabilizer parameters based on moth fire suppression optimization algorithm[J]. Power System Technology, 2020, 44(8):3038-3046.
[21]	LIN T K, LEE D Y. Composite multiscale cross-sample entropy analysis for long-term structural health monitoring of residential buildings[J]. Entropy, 2020, 23(1):60. doi: 10.3390/e23010060
[22]	刘强, 胡志强, 周宇, 等. 基于CEEMD和随机森林算法的短期风电功率预测[J]. 智慧电力, 2019, 47(6):71-76,94.
	LIU Qiang, HU Zhiqiang, ZHOU Yu, et al. Short-term wind power prediction based on CEEMD and random forest algorithm[J]. Smart Power, 2019, 47(6):71-76,94.

算法类型	相同迭代次数(300次)算法寻优结果
算法类型	f₁函数	f₂函数	f₃函数	f₄函数
改进MFO	0	2.53×10^-8	0	1.1×10^-274
MFO	0.005 25	4.02×10^-5	1.9×10^-5	0.009 831
PSO	109.863 5	1.97×10^-6	1.48×10^-23	4×10^-14
算法类型	相同收敛精度(10^-6)算法迭代次数(最大300次)
算法类型	f₁函数	f₂函数	f₃函数	f₄函数
改进MFO	7	98	11	7
MFO	未达到	未达到	67	94
PSO	未达到	未达到	未达到	299

模型	MAPE(%)	RMSE/kW
LSSVM	16.38	3.122
LSTM	12.28	1.718
Attention-LSTM	9.77	0.806
MFO-Attention-LSTM	6.88	0.289
改进MFO-Attention-LSTM	4.22	0.124