基于CNN-BiGRU-Attention的短期电力负荷预测

doi:10.11985/2024.01.037

摘要/Abstract

摘要：

针对目前电力负荷数据随机性强，影响因素复杂，传统单一预测模型精度低的问题，结合卷积神经网络(Convolutional neural network，CNN)、双向门控循环单元(Bi-directional gated recurrent unit，BiGRU)以及注意力机制(Attention)在短期电力负荷预测上的不同优点，提出一种基于CNN-BiGRU-Attention的混合预测模型。该方法首先通过CNN对历史负荷和气象数据进行初步特征提取，然后利用BiGRU进一步挖掘特征数据间时序关联，再引入注意力机制，对BiGRU输出状态给与不同权重，强化关键特征，最后完成负荷预测。试验结果表明，该模型的平均绝对百分比误差(Mean absolute percentage error，MAPE)、均方根误差(Root mean square error，RMSE)、判定系数(R-square，R²)分别为0.167%、0.057%、0.993，三项指标明显优于其他模型，具有更高的预测精度和稳定性，验证了模型在短期负荷预测中的优势。

关键词: 卷积神经网络, 双向门控循环单元, 注意力机制, 短期电力负荷预测, 混合预测模型

Abstract:

In the light of the problems of strong randomness of current power load data, complex influencing factors and low accuracy of traditional single forecasting model, combining the different advantages of convolutional neural network(CNN), bi-directional gated recurrent unit(BiGRU) and attention mechanism(Attention) in short-term load forecasting, a hybrid CNN-BiGRU-Attention based prediction model is proposed. First the initial features of historical load and meteorological data is extracted through CNN, then time series association of feature data is further excavated by BiGRU. Then attention mechanism is introduced to give different weights to the output status of BiGRU, strengthen key features. Finally, load prediction is completed. The experimental results show that the mean absolute percentage error(MAPE), root mean square error(RMSE) and R-square(R²) of the model are 0.167%, 0.057% and 0.993, respectively, and the three indicators are significantly better than other models, with higher prediction accuracy and stability. The advantages of this proposed model in short-term load forecasting are verified.

Key words: Convolutional neural network, bi-directional gated recurrent unit, attention mechanism, short-term power load forecasting, hybrid prediction model

中图分类号:

TP183

任爽, 杨凯, 商继财, 祁继明, 魏翔宇, 蔡永根. 基于CNN-BiGRU-Attention的短期电力负荷预测[J]. 电气工程学报, 2024, 19(1): 344-350.

REN Shuang, YANG Kai, SHANG Jicai, QI Jiming, WEI Xiangyu, CAI Yonggen. Short-term Power Load Forecasting Based on CNN-BiGRU-Attention[J]. Journal of Electrical Engineering, 2024, 19(1): 344-350.

图/表 12

图1

图2

图3

图4

图5

表1

表2

表3

图6

图7

图8

图9

参考文献 20

[1]	杨邓, 杨俊杰, 胡晨阳, 等. 基于改进LSSVM的短期电力负荷预测[J]. 电子测量技术, 2021, 44(18):47-53.
	YANG Deng, YANG Junjie, HU Chenyang, et al. Short term power load forecasting based on improved LSSVM[J]. Electronic Measurement Technology, 2021, 44(18):47-53.
[2]	陈纬楠, 胡志坚, 岳菁鹏, 等. 基于长短期记忆网络和LightGBM组合模型的短期负荷预测[J]. 电力系统自动化, 2021, 45(4):91-97.
	CHEN Weinan, HU Zhijian, YUE Jingpeng, et al. Short- term load prediction based on combined model of long and short-term memory network and light gradient boosting machine[J]. Automation of Electric Power Systems, 2021, 45(4):91-97.
[3]	姜东良, 李天昊, 刘文浩. 基于相似日和SAE-DBiLSTM模型的短期电力负荷预测[J]. 电气工程学报, 2022, 17(4):240-249.
	JIANG Dongliang, LI Tianhao, LIU Wenhao. Short term power load forecasting based on similar days and SAE-DBiLSTM model[J]. Journal of Electrical Engineering, 2022, 17(4):240-249.
[4]	董子晗. 基于混沌时间序列的地区电网负荷预测[J]. 电网与清洁能源, 2019, 35(5):38-41.
	DONG Zihan. Load forecasting of regional power grid based on the chaotic time series[J]. Power System and Clean Energy, 2019, 35(5):38-41.
[5]	DUDEK G. Pattern-based local linear regression models for short-term load forecasting[J]. Electric Power Systems Research, 2016,130:139-147.
[6]	赵峰, 孙波, 张承慧. 基于多变量相空间重构和卡尔曼滤波的冷热电联供系统负荷预测方法[J]. 中国电机工程学报, 2016, 36(2):399-406.
	ZHAO Feng, SUN Bo, ZHANG Chenghui. Cooling, heating and electric load forecasting method of CCHP system based on multivariate phase space reconstruction and Kalman filter[J]. Proceedings of the CSEE, 2016, 36(2):399-406.
[7]	陈培垠, 方彦军. 基于卡尔曼滤波预测节假日逐点增长率的电力系统短期负荷预测[J]. 武汉大学学报, 2020, 53(2):139-144.
	CHEN Peiyin, FANG Yanjun. Short term load forecasting of power system based on Kalman filter for predicting holiday point by point growth rate[J]. Engineering Journal of Wuhan University, 2020, 53(2):139-144.
[8]	EROSHENKO S A, POROSHIN V I, SENYUK M D, et al. Expert models for electric load forecasting of power system[C]// Young Researchers in Electrical & Electronic Engineering. IEEE,2017:1507-1513.
[9]	杨捷, 罗成臣, 张思路, 等. 相空间重构和SVR耦合的短期电力负荷预测[J]. 电测与仪表, 2020, 57(16):96-100.
	YANG Jie, LUO Chengchen, ZHANG Silu, et al. Short-term power load forecasting based on phase space reconstruction and SVR coupling model[J]. Electrical Measurement & Instrumentation, 2020, 57(16):96-100.
[10]	HOUIMLI R, ZMAMI M, BEN-SALHA O. Short-term electric load forecasting in Tunisia using artificial neural networks[J]. Energy Systems,2020:357-375.
[11]	YI N, JIA Y, YAN L M, et al. Mid-term human power load forecasting based on LSTM neural network[J]. Basic & Clinical Pharmacology & Toxicology, 2019,125:67-68.
[12]	朱凌建, 荀子涵, 王裕鑫, 等. 基于CNN-Bi LSTM的短期电力负荷预测[J]. 电网技术, 2021, 45(11):4532-4539.
	ZHU Lingjian, XUN Zihan, WANG Yuxin, et al. Short-term power load forecasting based on CNN-BiLSTM[J]. Power System Technology, 2021, 45(11):4532-4539.
[13]	党存禄, 杨海兰, 武文成. 基于LSTM和CatBoost组合模型的短期负荷预测[J]. 电气工程学报, 2021, 16(3):62-69.
	DANG Cunlu, YANG Hailan, WU Wencheng. Short-term load forecasting based on LSTM and CatBoost combined model[J]. Journal of Electrical Engineering, 2021, 16(3):62-69.
[14]	赵兵, 王增平, 纪维佳, 等. 基于注意力机制的CNN-GRU短期电力负荷预测方法[J]. 电网技术, 2019, 43(12):4370-4376.
	ZHAO Bing, WANG Zengping, JI Weijia, et al. A short-term power load forecasting method based on attention mechanism of CNN-GRU[J]. Power System Technology, 2019, 43(12):4370-4376.
[15]	方娜, 余俊杰, 李俊晓, 等. 基于CNN-BIGRU- ATTENTION的短期电力负荷预测[J]. 计算机仿真, 2022, 39(2):40-44,82.
	FANG Na, YU Junjie, LI Junxiao, et al. Short-term power load forecasting based on CNN-BIGRU-ATTENTION[J]. Computer Simulation, 2022, 39(2):40-44,82.
[16]	曾囿钧, 肖先勇, 徐方维, 等. 基于CNN-BiGRU-NN模型的短期负荷预测方法[J]. 中国电力, 2021, 54(9):17-23.
	ZENG Youjun, XIAO Xianyong, XU Fangwei, et al. Short term load forecasting method based on CNN-BiGRU-NN model[J]. Electric Power, 2021, 54(9):17-23.
[17]	HU R W, XIANG S J. CNN prediction based reversible data hiding[J]. IEEE Signal Processing Letters, 2021,28:464-468.
[18]	于军琪, 聂己开, 赵安军, 等. 基于特征挖掘的ARIMA-GRU短期电力负荷预测[J]. 电力系统及其自动化学报, 2022, 34(3):91-99.
	YU Junqi, NIE Jikai, ZHAO Anjun, et al. ARIMA-GRU short-term power load forecasting based on feature mining[J]. Proceedings of the CSU-EPSA, 2022, 34(3):91-99.
[19]	龚飘怡, 罗云峰, 方哲梅, 等. 基于Attention-BiLSTM- LSTM神经网络的短期电力负荷预测方法[J]. 计算机应用, 2021, 41(S1):81-86.
	GONG Piaoyi, LUO Yunfeng, FANG Zhemei, et al. Short-term power load forecasting method based on Attention-BiLSTM-LSTM neural network[J]. Journal of Computer Applications, 2021, 41(S1):81-86.
[20]	王金玉, 金宏哲, 王海生, 等. ISSA优化Attention双向LSTM的短期电力负荷预测[J]. 电力系统及其自动化学报, 2022, 34(5):111-117.
	WANG Jinyu, JIN Hongzhe, WANG Haisheng, et al. Short-term power load prediction of bidirectional LSTM with ISSA optimization Attention mechanism[J]. Proceedings of the CSU-EPSA, 2022, 34(5):111-117.

分类	范围
训练集	2012年1月1日—2013年12月31日
验证集	2014年1月1日—2014年6月30日
测试集	2014年7月1日—2014年12月31日

模型	超参数
模型	网络层数	神经元个数	迭代次数	批量大小	学习率
GRU	2	均为10	100	16	0.01
BiGRU	2	均为10	100	16	0.01
BiGRU-Attention	2	均为10	100	128	0.01
CNN-BiGRU-Attention	CNN：2	CNN1：32	150	128	0.01
	CNN：2	CNN2：64
	BiGRU：2	均为20

模型	MAPE(%)	RMSE(%)	R²
GRU	0.476	0.159	0.949
BiGRU	0.391	0.124	0.969
BiGRU-Attention	0.364	0.091	0.983
CNN-BiGRU-Attention	0.167	0.057	0.993