Short-term new energy power prediction based on TL-LSTM

doi:10.3969/j.issn.2097-0706.2023.01.005

Abstract

Abstract:

New energy power prediction is the key to realize accurate situation awareness of active distribution network.To deal with the uncertainty and volatility brought by new energy power generation，a transfer learning and long short-term memory（TL-LSTM）-based power prediction method is proposed.First，the k-shape clustering algorithm is used to cluster the time-series data provided by new energy sources in different regions，while each cluster can generate several training models.Then，the clusters closest to the target sequence are selected as auxiliary data clusters using shape-based distance （SBD）metrics for migration learning.The training of TL-LSTM models is completed with the pre-trained models corresponding to the auxiliary data clusters，and the difference treatment is used in all the model training processes to avoid the prediction lags.Finally，the effectiveness of the proposed prediction method is verified by a wind farm and a photovoltaic power plant in China as typical examples.The results show that the method improves the accuracy of short-term prediction on new energy units'outputs，and is widely applicable to new energy power prediction under small sample size.

Key words: power prediction for new energy, active distribution network, situation awareness, transfer learning, LSTM, k-shape algorithm, few-short learning, situational wareness

CLC Number:

TK01

ZHENG Zhen, ZHU Feng, MA Xiaoli, TIAN Shuxin, JIANG Haozhe. Short-term new energy power prediction based on TL-LSTM[J]. Integrated Intelligent Energy, 2023, 45(1): 41-48.

Figures/Tables 13

Fig.1

Fig.2

Fig.3

Fig.4

Fig.5

Fig.6

Fig.7

Table 1

Table 2

Table 3

Table 4

Fig.8

Fig.9

References 23

[1]	黄伟, 刘琦, 杨舒文, 等. 基于主动配电系统供电能力的安全态势感知方法[J]. 电力自动化设备, 2017, 37(8):74-80.
	HUANG Wei, LIU Qi, YANG Shuwen, et al. Security situational awareness approach based on power supply capacity of active distribution systems[J]. Electric Power Automation Equipment, 2017, 37(8):74-80.
[2]	王守相, 葛磊蛟. 主动配电系统运行与控制关键技术[J]. 电力建设, 2015, 36(1):85-90. doi: 10.3969/j.issn.1000－7229.2015.01.013
	WANG Shouxiang, GE Leijiao. Key technology of operation and control of active distribution system[J]. Electric Power Construction, 2015, 36(1):85-90. doi: 10.3969/j.issn.1000－7229.2015.01.013
[3]	王守相, 梁栋, 葛磊蛟. 智能配电网态势感知和态势利导关键技术[J]. 电力系统自动化, 2016, 40(12):2-8.
	WANG Shouxiang, LIANG Dong, GE Leijiao, et al. Smart distribution network situational awareness and situational leverage key technologies[J]. Automation of Electric Power Systems, 2016, 40(12):2-8.
[4]	葛磊蛟, 李元良, 陈艳波, 等. 智能配电网态势感知关键技术及实施效果评价[J]. 高电压技术, 2021, 47(7):2269-2280.
	GE Leijiao, LI Yuanliang, CHEN Yanbo, et al. Key technologies of situation awareness and implementation effectiveness evaluation in smart distribution network[J]. High Voltage Engineering, 2021, 47(7):2269-2280.
[5]	GEP B, DA P. Distribution of residual autocorrelations in autoregressive-integrated moving average time series models[J]. Journal of the American statistical Association, 1970, 65(332):1509-1526. doi: 10.1080/01621459.1970.10481180
[6]	HOCHREITER S, SCHMIDHUBER J. Long short-term memory[J]. Neural Computation, 1997, 9:1735-1780. doi: 10.1162/neco.1997.9.8.1735 pmid: 9377276
[7]	CHO K, MERRIËNBOER B V, GULCEHRE C, et al. Learning phrase representations using RNN encoder-decoder for statistical machine translation[C]// The Association for Computational Linguistics,Conference on Empirical Methods in Natural Language Processing, 2014.
[8]	王粟, 江鑫, 曾亮, 等. 基于VMD-DESN-MSGP模型的超短期光伏功率预测[J]. 电网技术, 2020, 44(3):917-926.
	WANG Su, JIANG Xin, ZENG Liang, et al. Ultra-short-term photovoltaic power prediction based on VMD-DESN-MSGP model[J]. Power System Technology, 2020, 44(3):917-926.
[9]	周楠, 徐潇源, 严正, 等. 基于宽度学习系统的光伏发电功率超短期预测[J]. 电力系统自动化, 2021, 45(1):55-64.
	ZHOU Nan, XU Xiaoyan, YAN Zheng, et al. Ultra-short-term prediction of photovoltaic power based on width learning system[J]. Automation of Electric Power Systems, 2021, 45(1):55-64.
[10]	游坤奇, 熊殷, 贾永青, 等. 基于PCC-RBF网络的风电功率短期预测方法[J]. 电机与控制应用, 2021, 48(1):41-45,104.
	YOU Kunqi, XIONG Yin, JIA Yongqing, et al. Short-term wind power forecast method based on pearson correlation coefficient and RBF network[J]. Electric Machines and Control Application, 2021, 48(1):41-45, 104.
[11]	杨晶显, 张帅, 刘继春, 等. 基于VMD和双重注意力机制LSTM的短期光伏功率预测[J]. 电力系统自动化, 2021, 45(3):174-182.
	YANG Jingxian, ZHANG Shuai, LIU Jichun, et al. Short-term PV power prediction based on VMD and dual attention mechanism LSTM[J]. Automation of Electric Power Systems, 2021, 45(3):174-182.
[12]	吉锌格, 李慧, 刘思嘉, 等. 基于MIE-LSTM的短期光伏功率预测[J]. 电力系统保护与控制, 2020, 48(7):50-57.
	JI Xinge, LI Hui, LIU Sijia, et al. Short-term photovoltaic power forecasting based on MIE-LSTM[J]. Power System Protection and Control, 2020, 48(7):50-57.
[13]	曹力, 潘巧波, 王明宇, 等. 基于混合核函数支持向量机的风电机组发电机温度预警方法[J]. 华电技术, 2020, 42(5):43-49.
	CAO Li, PAN Qiaobo, WANG Mingyu, et al. Early warning method for wind turbine generator temperature based on HK-SVM[J]. Huadian Technology, 2020, 42(5):43-49.
[14]	杨明明. 基于卷积神经网络的机舱风速修正[J]. 华电技术, 2021, 43(5):75-79.
	YANG Mingming. Wind speed correction for wind turbine based on convolutional neural network[J]. Huadian Technology, 2021, 43(5):75-79.
[15]	韩义, 张奇月, 王研凯, 等. 基于BP神经网络的300 MW循环流化床机组出力预测[J]. 华电技术, 2020, 42(12):1-6.
	HAN Yi, ZHANG Qiyue, WANG Yankai, et al. Performance prediction on a 300 MW CFB based on BP neural network[J]. Huadian Technology, 2020, 42(12):1-6.
[16]	杨毅, 范栋琛, 殷浩然, 等. 基于深度-迁移学习的输电线路故障选相模型及其可迁移性研究[J]. 电力自动化设备, 2020, 40(10):165-172.
	YANG Yi, FAN Dongchen, YIN Haoran, et al. Deep-migration learning based transmission line fault phase selection model and its migrability study[J]. Electric Power Automation Equipment, 2020, 40(10):165-172.
[17]	孙晓燕, 李家钊, 曾博, 等. 基于特征迁移学习的综合能源系统小样本日前电力负荷预测[J]. 控制理论与应用, 2021, 38(1):63-72.
	SUN Xiaoyan, LI Jiazhao, ZENG Bo, et al. Small-sample day-ahead power load forecasting of integrated energy system based on feature transfer learning[J]. Control Theory and Applications, 2021, 38(1):63-72.
[18]	王伟胜, 王铮, 董存, 等. 中国短期风电功率预测技术现状与误差分析[J]. 电力系统自动化, 2021, 45(1):17-27.
	WANG Weisheng, WANG Zheng, DONG Cun, et al. Current status and error analysis of short-term wind power forecasting technology in China[J]. Automation of Electric Power System, 2021, 45(1):17-27.
[19]	LIU F, TING K, ZHOU Z, et al. Isolation forest[C]// Institute of Electrical and Electronics Engineers.Eighth IEEE International Conference on Data Mining, 2008:413-422.
[20]	PAPARRIZOS J, LUIS G. K-Shape:Efficient and accurate clustering of time series[J]. Sigmod Record, 2016, 45:69-76.
[21]	ELMAN J. Finding structure in time[J]. Cognitive Science, 1990, 14(2):179-211. doi: 10.1207/s15516709cog1402_1
[22]	余光正, 陆柳, 汤波, 等. 考虑转折性天气的海上风电功率超短期分段预测方法研究[J]. 中国电机工程学报, 2022, 42(13):4859-4870.
	YU Guangzheng, LU Liu, TANG Bo, et al. Research on ultra-short-term subsection forecasting method of offshore wind power considering transitional weather[J]. Proceedings of the CSEE, 2022, 42(13):4859-4870.
[23]	YAO Q, SONG D, CHEN H, et al. A dual-stage attention-based recurrent neural network for time series prediction[C]// IJCAI Inc.International Joint Conference on Artificial Intelligence, 2017:2627-2623.

编号	SBD	编号	SBD
聚类簇1	0.994	聚类簇5	0.666
聚类簇2	0.640	聚类簇6	0.398
聚类簇3	0.693	聚类簇7	0.566
聚类簇4	0.693	聚类簇8	0.997

编号	SBD	编号	SBD
聚类簇1	0.638	聚类簇5	0.674
聚类簇2	0.627	聚类簇6	0.585
聚类簇3	0.617	聚类簇7	0.577
聚类簇4	0.619	聚类簇8	0.532

预测模型	δ_RMSE/kW	δ_MAE/kW	R²
TL-LSTM	1.580	1.235	0.910 0
LSTM	2.385	1.595	0.819 4
GRU	2.296	1.668	0.810 0
DA-LSTM	1.756	1.439	0.888 0

预测模型	δ_RMSE/kW	δ_MAE/kW	R²
TL-LSTM	3.239	1.853	0.728
LSTM	3.855	2.350	0.615
GRU	3.817	2.173	0.622
DA-LSTM	3.644	2.693	0.656