基于多模式特征筛选和改进DeepAR的风电功率多步预测

李新利; 罗政龙; 张奎; 杨国田

doi:10.19725/j.cnki.1007-2322.2024.0167

欢迎来到能源电力期刊网！

English Version 期刊登录

您当前的位置：

首页 >

文章列表页 >

基于多模式特征筛选和改进DeepAR的风电功率多步预测

新能源电力系统 | 更新时间：2026-10-19

- 基于多模式特征筛选和改进DeepAR的风电功率多步预测
- Multi-step Prediction of Wind Power Based on Multi-mode Feature Screening and Enhanced DeepAR
- 现代电力 2026年43卷第5期页码：1012-1022
- 作者机构：
  
  华北电力大学控制与计算机工程学院, 北京市昌平区,102206
- 作者简介：
  
  李新利(1973)，女，博士，副教授，通信作者，研究方向为综合能源系统、模式识别与智能系统、图像处理，E-mail：lixinli@ncepu.edu.cn
  罗政龙(2000)，男，硕士研究生，研究方向为综合能源系统调度与优化，E-mail：luozhenglong_ncepu@163.com
  张奎(1999)，男，硕士研究生，研究方向为综合能源系统调度与优化，E-mail：zhangkui_ncepu@163.com
  杨国田(1962)，男，博士，教授，研究方向为模式识别与智能系统、机器视觉、嵌入式系统研发，E-mail：ygt@ncepu.edu.cn
- 基金信息：
- DOI：10.19725/j.cnki.1007-2322.2024.0167
  中图分类号： 5
- 纸质出版：2026
- 稿件说明：
移动端阅览
李新利, 罗政龙, 张奎, 等. 基于多模式特征筛选和改进DeepAR的风电功率多步预测[J]. 现代电力, 2026,43(5):1012-1022.

LI Xinli, LUO Zhenglong, ZHANG Kui, et al. Multi-step Prediction of Wind Power Based on Multi-mode Feature Screening and Enhanced DeepAR[J]. 2026, 43(5): 1012-1022.
李新利, 罗政龙, 张奎, 等. 基于多模式特征筛选和改进DeepAR的风电功率多步预测[J]. 现代电力, 2026,43(5):1012-1022. DOI： 10.19725/j.cnki.1007-2322.2024.0167.

LI Xinli, LUO Zhenglong, ZHANG Kui, et al. Multi-step Prediction of Wind Power Based on Multi-mode Feature Screening and Enhanced DeepAR[J]. 2026, 43(5): 1012-1022. DOI： 10.19725/j.cnki.1007-2322.2024.0167.

摘要

随着新能源机组装机容量的持续增加，实现风电出力的精准预测已成为大规模风机并网的关键。针对风电功率多步预测问题，提出一种基于贝叶斯–深度自回归(Bayes-deep autoregressive，Bayes-DeepAR)的风电功率预测模型。首先，构建一种多模式特征筛选策略，融合皮尔逊相关系数、F检验和平均影响值算法的特点，筛选出对风电功率影响较大的特征变量集。其次，提出基于嵌入层、长短期记忆网络层和概率分布层的DeepAR预测模型，采用贝叶斯方法优化模型超参数，实现多步预测。最后，基于预测结果，从横向误差、纵向误差以及误差累积指标综合分析时间序列多步预测的误差特征。与其他预测模型进行对比，仿真结果表明所提多步预测模型不仅能实现风电功率的点预测和概率区间预测，而且能提升预测精度。

Abstract

With the increasing installed capacity of new energy units

accurate prediction of wind power output has become crucial for realizing large-scale wind turbine grid-connection. To address the issue of multi-step prediction of wind power

a model of wind power prediction is proposed based on Bayes-DeepAR. Firstly

a multi-mode feature screening strategy is constructed

integrating the Pearson correlation coefficient

F-test and mean impact value algorithm to screen out the feature variable set that has a greater impact on wind power. Secondly

a DeepAR prediction model is proposed grounded in embedded-layer

long-short-term memory network layer and Gaussian distribution layer. The Bayes approach is utilized to optimize the hyperparameters of the model

aiming to achieve multi-step prediction. Finally

based on the prediction results

a comprehensive analysis on the error characteristics of time series multi-step prediction is conducted

containing indexes such as the horizontal error

vertical error and error accumulation. Compared to other prediction models

the simulation results demonstrate that the proposed multi-step prediction model not only achieves point prediction and probability interval prediction of wind power

but also improves the prediction accuracy.

关键词

Keywords

references

CHEN Feng, YU Yi, XU Jingyou, et al . Prediction method of wind power output based on a Bayes-LSTM network[J ] . Power System Protection and Control, 2023, 51(06): 170−178(in Chinese).

XU Jin. Overview of China's energy and electricity development in 2023[J]. Energy, 2024(03): 29−35(in Chinese).

JIANG Yu, CHEN Xingying, YU Kun, et al . Short-term wind power forecasting using hybrid method based on enhanced boosting algorithm[J ] . Journal of Modern Power Systems and Clean Energy, 2017, 5(1): 126−133.

YANG Guohua, QI Xin, JIA Rui, et al . Short-term wind power forecast based on CNNLSTM-GRU model integrated with CEEMD-SE algorithm[J ] . Electric Power, 2024, 57(02): 55−61(in Chinese).

WU Yong, WANG Bing, CHEN Yuquan, et al . Application of deep learning model integrating refined meteorological factors and physical constraints in short-term wind power prediction[J ] . Power System Technology, 2024, 48(04): 1455−1468(in Chinese).

LU Jiping, ZENG Yanting, YU Hua, et al . Ultra-short term wind power multi-step forecasting based on improved AWNN[J ] . Acta Energiae Solaris Sinica, 2021, 42(01): 166−173(in Chinese).

YANG Deyou, GAO Ziang, LI Yinxuan. Interval prediction of wind power based on bivariate empirical mode decomposition and least squares support vector machine[J]. Electric Power Construction, 2019, 40(05): 118−127(in Chinese).

XU Yan, XIANG Yifeng, MA Tianxiang. Short-term power load forecasting method based on EMD-CNN-LSTM hybrid model[J]. Journal of North China Electric Power University(Natural Science Edition), 2022, 49(02): 81−89(in Chinese).

CUI Mingyong, DONG Wentao, LU Zhigang. CEEMDAN-CNN-LSTM short-term wind power prediction based on density clustering[J]. Modern Electric Power, 2024, 41(04): 631−641(in Chinese).

LIU Xinyu, PU Xinyu, JI Fang, et al . Short-term wind power prediction of a VMD-GRU based on Bayesian optimization[J ] . Power System Protection and Control, 2023, 51(21): 158−165(in Chinese).

XIAO Liexi, ZHANG Yu, ZHOU Hui, et al . Ultra short term wind power prediction based on IAOA-VMD-LSTM[J ] . Acta Energiae Solaris Sinica, 2023, 44(11): 239−246(in Chi nese).

LIANG Changhou, LONG Hua, LI Shuai, et al . Short-term wind power prediction based on MIC-VMD-GWO-LSTM[J ] . Modern Electronics Technique, 2023, 46(22): 115−120(in Chinese).

PANG Bowen, DING Yueming, DU Shanhui, et al . Short-term forecasting of wind power output based on CEEMDAN-BO-LSTNet[J ] . Electrical Measurement Instrumentation, 2023, 60(09): 109−116+170(in Chinese).

ZHAO Dongmei, DU Gang, LIU Xin, et al . Wind power combination prediction model based on time series decomposition and machine learning[J ] . Modern Electric Power, 2022, 39(01): 9−19(in Chinese).

ZHANG Yali, WANG Cong, ZHANG Hongli, et al . Multi-step prediction of ultra-short-term wind power based onnon-stationary transformer[J ] . Smart Power, 2024, 52(01): 108−115(in Chinese).

LIU Yu, ZENG Deliang, LIU Jizhen, et al . Wavelet packet transform and the least squares support vector machine in research of short-term wind speed multistep prediction and information granulation prediction[J ] . Acta Energiae Solaris Sinica, 2014, 35(02): 214−220(in Chinese).

WU Jian, XIANG Song, YAN Cheng, et al . Multi-step prediction of super-short-term wind power based on singular spectrum analysis[J ] . Renewable Energy Resources, 2021, 39(11): 1548−1555(in Chinese).

WANG Lei, HE Yigang. M2STAN: multi-modal multi-task spatiotemporal attention network for multi-location ultra-short-term wind power multi-step predictions[J]. Applied Energy, 2022(4): 324.

WANG Yongsheng, ZHANG Zhe, LIU Limin, et al . Short-term wind power prediction based on improved entropy weight method and SECEEMD[J ] . Science Technology and Engineering, 2023, 23(27): 11688−11697(in Chinese).

YANG Kewen, SUN Yingyun. Probability prediction of ultra-short-term wind power based on temporal pattern attention[J]. Modern Electric Power, 2023, 40(06): 906−913(in Chinese).

PEARSON K. Notes on the history of correlation[J]. Biometrika, 1920, 13(1): 25−45.

FLUNKERT V, SALINAS D, GASTHAUS J. DeepAR: probabilistic forecasting with autoregressive recurrent networks[J]. International Journal of Forecasting, 2020, 36(3): 1181−1191.

CUI Ro ngfeng, MA Hai, GUO Chengpeng, et al . Prediction of missile aerodynamic data based on gradient boosting under Bayesian hyperparametric optimization[J ] . Aeronautical Science Technology, 2023, 34(07): 22−28(in Chinese).

LI Yaru, ZHANG Yulai, WANG Jiachen. Survey on Bayesian optimization methods for hyper-parameter tuning[J]. Computer Science, 2022, 49(supplement 1): 86−92(in Chinese).

LETHAM B, KARRER B, OTTONI G, et al . Constrained Bayesian optimization with noisy experiments[J ] . Bayesian Analysis, 2019, 14(02): 495−519.

LI Fengming, CHE Runqi, YANG Mao. Evaluation and analysis of wind power ultra-short-term multi-step forecast error[J]. Journal of Kunming University of Science and Technology (Natural Science Edition), 2024, 49(01): 106−114(in Chinese).

浏览量

280

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

考虑特征重组和BiGRU-Attention-XGBoost模型的超短期负荷功率预测

基于多源停电数据提示学习的电网轻量化停电感知模型

基于区块链与安全计算的电能质量扰动分析方法

一种基于特征映射与深度学习的虚假数据注入检测方法

基于生成对抗网络的自爆绝缘子检测模型设计