Informer Photovoltaic Power Generation Forecasting Based on Cycle Information Enhancement

doi:10.11930/j.issn.1004-9649.202306038

Abstract

Abstract:

Aiming at the problem that the photovoltaic power generation power prediction method is difficult to capture hierarchical time series information, which results in the limited improvement of prediction accuracy. The paper proposes a photovoltaic power generation prediction method based on the fusion of hierarchical time series information and Informer model. First, photovoltaic power sequence scalar projection, local time stamp and global time stamp are extracted to establish a prediction model embedding layer of periodic information enhancement. Then, through the probability sparse self-attention of the Informer coding layer, the important connection between the photovoltaic power generation and the characteristic variables is actively screened. The convolutional layer and pooling layer are used to optimize the model variable dimensions and network parameters for self-attention distillation. Finally, one-step prediction of single-sequence and long-sequence power generation is realized through the generative mechanism of the decoding layer. Through simulation verification, the proposed model has higher prediction accuracy and can make long-term predictions of photovoltaic power generation.

Key words: photovoltaic power generation forecasting, cycle information enhancement, self-attention, long sequence time-series forecasting

Qian ZHANG, Fei MENG, Tao LI, Yong YANG, Lu BAI. Informer Photovoltaic Power Generation Forecasting Based on Cycle Information Enhancement[J]. Electric Power, 2023, 56(10): 186-193.

Add to citation manager EndNote|Ris|BibTeX

URL: https://www.electricpower.com.cn/EN/10.11930/j.issn.1004-9649.202306038

https://www.electricpower.com.cn/EN/Y2023/V56/I10/186

Figures/Tables 10

Fig.1 Informer network model structure

Fig.2 Training process of informer photovoltaic power generation forecasting model

Table 1 The fluctuation range of each variable data

变量		变化区间
时		[1, 24]
月		[1, 12]
周		[1, 7]
日		[1, 31]
全射辐照度/(W·m^–2)		[1, 1206]
直射辐照度/(W·m^–2)		[37, 938]
散射辐照度/(W·m^–2)		[1, 1013]
光伏组件温度/℃		[6, 34.4]
组件环境温度/℃		[4.6, 37.7]
组件环境湿度/(g·m^–3)		[14.9, 97.7]
光伏发电功率/MW		[0.07, 108.91]

Fig.3 Forecast effect of photovoltaic power generation

Fig.4 Comparison of models forecasting effect

Table 2 Comparison of forecast errors of different models

预测方法	e_MAPE/%	e_RMSE/MW
RNN	38.51	8.03
LSTM	15.93	6.25
Transformer	7.28	5.19
Informer	5.13	2.75

Table 3 Long sequence time-series forecasting effect of model 单位：MW

Informer预测模型	5月2日	5月3日	5月4日
1	3.16	4.62	5.29
2	—	3.29	4.84
3	—	—	3.52

Fig.5 Forecast effect in the next 3 days under current conditions

Fig.6 Forecast effect in the next 2 days under current conditions

Fig.7 Forecast effect in the next 1 day under current conditions

References 25

1	邱伟强, 王茂春, 林振智, 等. “双碳”目标下面向新能源消纳场景的共享储能综合评价[J]. 电力自动化设备, 2021, 41 (10): 244- 255.
	QIU Weiqiang, WANG Maochun, LIN Zhenzhi, et al. Comprehensive evaluation of shared energy storage towards new energy accommodation scenario under targets of carbon emission peak and carbon neutrality[J]. Electric Power Automation Equipment, 2021, 41 (10): 244- 255.
2	赵亮, 刘友波, 余莉娜, 等. 基于深度信念网络的光伏电站短期发电量预测[J]. 电力系统保护与控制, 2019, 47 (18): 11- 19.
	ZHAO Liang, LIU Youbo, YU Lina, et al. Short-term power generation forecast of PV power station based on deep belief network[J]. Power System Protection and Control, 2019, 47 (18): 11- 19.
3	HAMDI H, BEN REGAYA C, ZAAFOURI A. A sliding-neural network control of induction-motor-pump supplied by photovoltaic generator[J]. Protection and Control of Modern Power Systems, 2019, 5 (1): 1- 17.
4	李正明, 梁彩霞, 王满商. 基于PSO-DBN神经网络的光伏短期发电出力预测[J]. 电力系统保护与控制, 2020, 48 (8): 149- 154.
	LI Zhengming, LIANG Caixia, WANG Manshang. Short-term power generation output prediction based on a PSO-DBN neural network[J]. Power System Protection and Control, 2020, 48 (8): 149- 154.
5	张倩, 马愿, 李国丽, 等. 频域分解和深度学习算法在短期负荷及光伏功率预测中的应用[J]. 中国电机工程学报, 2019, 39 (8): 2221- 2230.
	ZHANG Qian, MA Yuan, LI Guoli, et al. Applications of frequency domain decomposition and deep learning algorithms in short-term load and photovoltaic power forecasting[J]. Proceedings of the CSEE, 2019, 39 (8): 2221- 2230.
6	时珉, 王强, 王铁强, 等. 基于特征筛选与ANFIS-PSO的分布式光伏发电功率预测方法研究[J]. 可再生能源, 2019, 37 (7): 989- 994.
	SHI Min, WANG Qiang, WANG Tieqiang, et al. Short-term distributed photovoltaic forecasting based on ANFIS-PSO and feature selection[J]. Renewable Energy Resources, 2019, 37 (7): 989- 994.
7	刘昳娟, 陈云龙, 刘继彦, 等. 基于集成学习的分布式光伏发电功率日前预测[J]. 中国电力, 2022, 55 (9): 38- 45.
	LIU Yijuan, CHEN Yunlong, LIU Jiyan, et al. Day-ahead prediction of distributed photovoltaic power generation based on ensemble learning[J]. Electric Power, 2022, 55 (9): 38- 45.
8	王俊杰, 毕利, 张凯, 等. 基于多特征融合和XGBoost-LightGBM-ConvLSTM的短期光伏发电量预测[J]. 太阳能学报, 2023, 44 (7): 168- 174.
	WANG Junjie, BI Li, ZHANG Kai, et al. Short-term photovoltaic power generation prediction based on multi-feature fusion and XGBoost-LightGBM-ConvLSTM[J]. Acta Energiae Solaris Sinica, 2023, 44 (7): 168- 174.
9	范士雄, 刘幸蔚, 於益军, 等. 基于多源数据和模型融合的超短期母线负荷预测方法[J]. 电网技术, 2021, 45 (1): 243- 250.
	FAN Shixiong, LIU Xingwei, YU Yijun, et al. Ultra-short-term bus load forecasting method based on multi-source data and model fusion[J]. Power System Technology, 2021, 45 (1): 243- 250.
10	唐旭辰, 潮铸, 段秦尉, 等. 基于分层测量数据的高压变电站概率负荷预测方法[J]. 中国电力, 2023, 56 (8): 143- 150.
	TANG Xuchen, CHAO Zhu, DUAN Qinwei, et al. Probabilistic load forecasting method of high voltage substation based on hierarchical measurement data[J]. Electric Power, 2023, 56 (8): 143- 150.
11	张立峰, 刘旭. 基于CNN-GRU神经网络的短期负荷预测[J]. 电力科学与工程, 2020, 36 (11): 53- 57.
	ZHANG Lifeng, LIU Xu. Short-term load forecasting based on CNN-GRU neural network[J]. Electric Power Science and Engineering, 2020, 36 (11): 53- 57.
12	崔佳豪, 毕利. 基于混合神经网络的光伏电量预测模型的研究[J]. 电力系统保护与控制, 2021, 49 (13): 142- 149.
	CUI Jiahao, BI Li. Research on photovoltaic power forecasting model based on hybrid neural network[J]. Power System Protection and Control, 2021, 49 (13): 142- 149.
13	王志远, 王守相, 陈海文, 等. 考虑空间相关性采用LSTM神经网络的光伏出力短期预测方法[J]. 电力系统及其自动化学报, 2020, 32 (5): 78- 85.
	WANG Zhiyuan, WANG Shouxiang, CHEN Haiwen, et al. Short-term photovoltaic output forecasting method using LSTM neural network with consideration of spatial correlation[J]. Proceedings of the CSU-EPSA, 2020, 32 (5): 78- 85.
14	庞传军, 张波, 余建明. 基于LSTM循环神经网络的短期电力负荷预测[J]. 电力工程技术, 2021, 40 (1): 175- 180, 194.
	PANG Chuanjun, ZHANG Bo, YU Jianming. Short-term power load forecasting based on LSTM recurrent neural network[J]. Electric Power Engineering Technology, 2021, 40 (1): 175- 180, 194.
15	陈振宇, 刘金波, 李晨, 等. 基于LSTM与XGBoost组合模型的超短期电力负荷预测[J]. 电网技术, 2020, 44 (2): 614- 620.
	CHEN Zhenyu, LIU Jinbo, LI Chen, et al. Ultra short-term power load forecasting based on combined LSTM-XGBoost model[J]. Power System Technology, 2020, 44 (2): 614- 620.
16	姚程文, 杨苹, 刘泽健. 基于CNN-GRU混合神经网络的负荷预测方法[J]. 电网技术, 2020, 44 (9): 3416- 3424.
	YAO Chengwen, YANG Ping, LIU Zejian. Load forecasting method based on CNN-GRU hybrid neural network[J]. Power System Technology, 2020, 44 (9): 3416- 3424.
17	王文广, 赵文杰. 基于GRU神经网络的燃煤电站NO_x排放预测模型 [J]. 华北电力大学学报(自然科学版), 2020, 47 (1): 96- 103.
	WANG Wenguang, ZHAO Wenjie. NO_x emission prediction model based on GRU neural network in coal-fired power station [J]. Journal of North China Electric Power University (Natural Science Edition), 2020, 47 (1): 96- 103.
18	谢谦, 董立红, 厍向阳. 基于Attention-GRU的短期电价预测[J]. 电力系统保护与控制, 2020, 48 (23): 154- 160.
	XIE Qian, DONG Lihong, SHE Xiangyang. Short-term electricity price forecasting based on Attention-GRU[J]. Power System Protection and Control, 2020, 48 (23): 154- 160.
19	宋绍剑, 李博涵. 基于LSTM网络的光伏发电功率短期预测方法的研究[J]. 可再生能源, 2021, 39 (5): 594- 602.
	SONG Shaojian, LI Bohan. Short-term forecasting method of photovoltaic power based on LSTM[J]. Renewable Energy Resources, 2021, 39 (5): 594- 602.
20	张雲钦, 程起泽, 蒋文杰, 等. 基于EMD-PCA-LSTM的光伏功率预测模型[J]. 太阳能学报, 2021, 42 (9): 62- 69.
	ZHANG Yunqin, CHENG Qize, JIANG Wenjie, et al. Photovoltaic power prediction model based on EMD-PCA-LSTM[J]. Acta Energiae Solaris Sinica, 2021, 42 (9): 62- 69.
21	周恒俊, 王璇, 王志远, 等. 基于MIPCA与GRU网络的光伏出力短期预测方法[J]. 电力系统及其自动化学报, 2020, 32 (9): 55- 62.
	ZHOU Hengjun, WANG Xuan, WANG Zhiyuan, et al. Short-term photovoltaic output prediction method based on MIPCA and GRU network[J]. Proceedings of the CSU-EPSA, 2020, 32 (9): 55- 62.
22	刘倩, 胡强, 杨凌帆, 等. 基于时间序列的深度学习光伏发电模型研究[J]. 电力系统保护与控制, 2021, 49 (19): 87- 98.
	LIU Qian, HU Qiang, YANG Lingfan, et al. Deep learning photovoltaic power generation model based on time series[J]. Power System Protection and Control, 2021, 49 (19): 87- 98.
23	王晨阳, 段倩倩, 周凯, 等. 基于遗传算法优化卷积长短记忆混合神经网络模型的光伏发电功率预测[J]. 物理学报, 2020, 69 (10): 100701.
	WANG Chenyang, DUAN Qianqian, ZHOU Kai, et al. A hybrid model for photovoltaic power prediction of both convolutional and long short-term memory neural networks optimized by genetic algorithm[J]. Acta Physica Sinica, 2020, 69 (10): 100701.
24	ZHOU H Y, ZHANG S H, PENG J Q, et al. Informer: beyond efficient transformer for long sequence time-series forecasting[C]// Proceedings of the AAAI Conference on Artificial Intelligence. New York, USA. AAAI, 2021: 11106–11115.
25	高发骏, 王怀远, 党然. 基于Transformer的暂态稳定评估模型的可解释性分析与模型更新研究[J]. 电力系统保护与控制, 2023, 51 (17): 15- 25.
	GAO Fajun, WANG Huaiyuan, DANG Ran. Interpretability analysis and model update research of a transient stability assessment model based on Transformer[J]. Power System Protection and Control, 2023, 51 (17): 15- 25.