为有效提取综合能源系统（IES）中数据的相关特征,并识别多元负荷预测任务中的重要特征,同时提高多元负荷的预测精度和速度,提出一种基于改进的卷积神经网络（CNN）和注意力机制的综合能源系统多元负荷短期预测模型,实现对多元负荷的双重特征识别和精确预测。首先,对数据进行相关性分析（ICEEMDAN）,筛选出与多元负荷相关性强的气象特征后,对电负荷、冷负荷和热负荷数据进行改进的自适应噪声完备集合经验模态分解,高效地将复杂的多元负荷时间序列分解为K个固有模态（IMFs）,并对IMFs进行相关性分析,筛选出关键IMFs。其次,利用CNN进行首次特征识别,提取数据的局部重要特征,再利用门控循环单元（GRU）捕捉时间序列数据的动态变化。然后,引入注意力机制实现双重特征识别,增强模型对关键特征的关注度,提高预测的准确性。最后,使用美国亚利桑那州立大学Tempe校区的多元负荷数据进行试验测试,同时采用相关模型进行对比分析。结果表明,所提方法的电、冷、热负荷平均绝对百分比误差分别为5.1%、4.23%及3.68%,相比其他模型具有更高的预测精度和更快的运算速率。
Abstract
This paper proposes a method in order to effectively extract the relevant features of the data in the integrated energy system（IES）,identify the important features in the multi-load forecasting task,and improve the prediction accuracy and speed of multi-load forecasting. In this paper,a short-term prediction model of multiple loads in an integrated energy system based on an improved convolutional neural network（CNN） and an attention mechanism is proposed,which realizes dual feature recognition and accurate prediction of multiple loads. Firstly,the correlation analysis of the data is carried out to screen out the meteorological characteristics with strong correlation with multiple loads,and then improved complete ensemble empirical mode decomposition with adaptive noise （ICEEMDAN） is carried out on the electrical, cooling and heating load data, and the complex multi-load time series is efficiently decomposed into K intrinsic modes （IMFs）, and the correlation analysis of IMFs is carried out to screen out the key IMFs. Secondly,CNN is used for feature recognition,wherelocal important features of the data are extracted,and then the gated recurrent unit （GRU） is used to capture the dynamic changes of the time series data. Then,the attention mechanism is introduced to realize the dual feature recognition,enhance the attention of the model to the key features,and improve the accuracy of prediction. Finally,the multivariate load data of Arizona State University Tempe Campus were used for experimental testing,and comparative models were used for comparative analysis. The results show that the average absolute percentage errors of electrical,cooling and heating loads of the proposed method are 5.1%,4.23% and 3.68%,respectively,which has higher prediction accuracy and faster computational speed than other models.
关键词
综合能源系统 /
负荷预测 /
模态分解 /
卷积神经网络 /
门控循环单元 /
注意力机制 /
特征提取
Key words
integrated energy systems /
load forecasting /
modal decomposition /
convolutional neural networks /
gated recurrent unit /
attention mechanism /
feature extraction
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 窦真兰, 张春雁, 许一洲, 等. 基于多变量相空间重构和径向基函数神经网络的综合能源系统电冷热超短期负荷预测[J]. 电网技术, 2024, 48(1): 121-128.
DOU Z L, ZHANG C Y, XU Y Z, et al.Ultra-short-term load forecasting of electrical, cooling and heating for integrated energy system based on multivariate phase space reconstruction and radial basis function neural network[J]. Power system technology, 2024, 48(1): 121-128.
[2] 欧阳静, 杨吕, 尹康, 等. 基于ALIF-LSTM多任务学习的综合能源系统短期负荷预测[J]. 太阳能学报, 2022, 43(9): 499-507.
OU Y J, YANG L, YIN K, et al.Short-term load forecasting method for integrated energy system based on ALIF-LSTM and multi-task learning[J]. Acta energiae solaris sinica, 2022, 43(9): 499-507.
[3] 秦烁, 赵健, 徐剑, 等. 基于多任务学习和单任务学习组合模型的综合能源系统多元负荷预测[J]. 电网技术, 2024,48(4): 1510-1518.
QIN S, ZHAO J, XU J, et al.Multivariate-load forecasting of integrated energy system based on combined multi-task learning and single-task learning model[J]. Power system technology, 2024, 48(4): 1510-1522.
[4] 邢晓萱, 巩敦卫, 孙晓燕, 等. 基于重构误差和极端模式识别的综合能源系统短期负荷预测[J]. 中国电机工程学报, 2024, 44(9): 3476-3488.
XING X X, GONG D W, SUN X Y, et al.Short-term load forecasting of integrated energy system based on reconstruction error and extreme patterns recognition[J]. Proceedings of the CSEE, 2024, 44(9): 3476-3488.
[5] 王小明, 徐斌, 尹元亚, 等. 基于空间相关性与Stacking集成学习的风电功率预测方法[J]. 电力工程技术, 2024, 43(5): 224-232.
WANG X M, XU B, YIN Y Y, et al.Wind power prediction method based on spatial correlation and Stacking ensemble learning[J]. Electric power engineering technology, 2024, 43(5): 224-232.
[6] TELLE J S, UPADHAYA A, SCHNFELDT P, et al.Probabilistic net load forecasting framework for application in distributed integrated renewable energy systems[J]. Energy reports, 2024, 11: 2535-2553.
[7] LIU F, LIANG C.Short-term power load forecasting based on AC-BiLSTM model[J]. Energy reports, 2024, 11: 1570-1579.
[8] 薛东, 段立强, 高统彤, 等. 考虑多重特征与不确定性度量的综合能源系统负荷预测研究[J]. 太阳能学报, 2024, 45(7): 379-388.
XUE D, DUAN L Q, GAO T T, et al.Study of integrated energy system load forecasting considering multiple characteristics and uncertainty measures[J]. Acta energiae solaris sinica, 2024, 45(7): 379-388.
[9] TAN Z F, DE G, LI M, et al.Combined electricity-heat-cooling-gas load forecasting model for integrated energy system based on multi-task learning and least square support vector machine[J]. Journal of cleaner production, 2019, 248(14): 1190-252.
[10] 王琛, 王颖, 郑涛, 等. 基于ResNet-LSTM网络和注意力机制的综合能源系统多元负荷预测[J]. 电工技术学报, 2022, 37(7): 1789-1799.
WANG C, WANG Y, ZHENG T, et al.Multi-energy load forecasting in integrated energy system based on ResNet-LSTM network and attention mechanism[J]. Transactions of China Electrotechnical Society, 2022, 37(7): 1789-1799.
[11] 黄璜, 张安安. 基于分解算法与元学习结合的综合能源系统负荷预测[J]. 电力系统自动化, 2024, 48(10): 151-160.
HUANG H, ZHANG A A.Load dorecasting of integrated energy system based on combination of decomposition algorithms and meta-learning[J]. Automation of electric power systems, 2024, 48(10): 151-160.
[12] 李云松, 张智晟. 考虑综合需求响应的Transformer-图神经网络综合能源系统多元负荷短期预测[J]. 电工技术学报, 2024, 39(19): 6119-6128.
LI Y S, ZHANG Z S.Transformer based multi load short-term forecasting of integrated energy system considering integrated demand response[J]. Transactions of China Electrotechnical Society, 2024, 39(19): 6119-6128.
[13] 王东风, 刘婧, 黄宇, 等. 结合太阳辐射量计算与CNN-LSTM组合的光伏功率预测方法研究[J]. 太阳能学报, 2024, 45(2): 443-450.
WANG D F, LIU J, HUANG Y, et al.Photovoltaic power prediction method combinating solar radiation calculation and CNN-LSTM[J]. Acta energiae solaris sinica, 2024, 45(2): 443-450.
[14] 杨晨, 侯志强, 李新月, 等. 基于CNN-Transformer双模态特征融合的目标检测算法[J]. 光子学报, 2024, 53(3): 280-293.
YANG C, HOU Z Q, LI X Y, et al.Object detection algorithm based on CNN-Transformer dual modal deature dusion[J]. Acta photonica sinica, 2024, 53(3): 280-293.
[15] GAO X, WU S, HU W X.Lightweight image super-resolution via multi-branch aware CNN and efficient transformer[J]. Neural computing & applications, 2024, 36(10): 5285-5303.
[16] 陈以浩, 李明伟, 安小刚, 等. 基于混沌云量子蝙蝠CNN-GRU大坝变形智能预报方法研究[J]. 哈尔滨工程大学学报, 2024, 45(1): 110-118.
CHEN Y H, LI M W, AN X G, et al.Research on the intelligent prediction method of dam deformation based on chaotic cloud quantum bat CNN-GRU[J]. Journal of Harbin Engineering University, 2024, 45(1): 110-118.
[17] 付国忠, 杜华, 张志强, 等. 基于注意力机制和CNN-BiLSTM模型的滚动轴承剩余寿命预测[J]. 核动力工程, 2023, 44(S2): 33-38.
FU G Z, DU H, ZHAG Z Q, et al.Remaining useful life prediction of rolling bearings based on Attention Mechanism and CNN-BiLSTM[J]. Nuclear power engineering, 2023, 44(S2): 33-38.
[18] FANG W, ZHANG S, XU C.Improving prediction efficiency of Chinese stock index futures intraday price by VIX-Lasso-GRU model[J]. Expert systems with applications, 2024, 238: 0957-4174.
[19] YANG Y F, LIAO S M, LIU M B.Dynamic prediction of moving trajectory in pipe jacking: GRU-based deep learning framework[J]. Frontiers of structural and civil engineering, 2023, 17(7): 994-1010.
[20] 石庆兰, 束金阳, 李道亮, 等. 基于BiLSTM-GRU融合网络的稻虾养殖溶解氧含量预测[J]. 农业机械学报, 2023, 54(10): 364-370.
SHI Q L, SHU J Y, LI D L, et al.Dissolved oxygen prediction in rice and shrimp culture based on BiLSTM-GRU fusion neural networks[J]. Transactions of the CSAM, 2023, 54(10): 364-370.
[21] 冉启武, 石卓见, 刘阳, 等. 考虑复合指标优化模态分解和Stacking集成的综合能源系统多元负荷预测[J]. 电网技术, 2025, 49(3): 1098-1108.
RAN Q W, SHI Z J, LIU Y, et al.Stacking ensemble for multi-energy load forecasting in integrated energy systems with consideration of optimized modal decomposition based on composite indices[J]. Power system technology, 2025, 49(3): 1098-1108.
[22] 冯佳威, 王海鑫, 杨子豪, 等. 基于数字孪生与多模型融合的多元负荷短期预测[J]. 太阳能学报, 2024, 45(10): 97-106.
FENG J W, WANG H X, YANG Z H, et al.Short-term forecasting of multivariate load based on digital twin and multi-model fusion[J]. Acta energiae solaris sinica, 2024, 45(10): 97-106.
[23] LI K, DUAN P, CAO X, et al.A multi-energy load forecasting method based on complementary ensemble empirical model decomposition and composite evaluation factor reconstruction[J]. Applied energy, 2024, 365: 0306-2619.
PDF(3303 KB)
