基于气象辐照数据和引力搜索算法优化VMD-BiLSTM的光伏发电功率预测研究

马海洋; 成贵学; 王函韵; 赵晋斌

doi:10.19725/j.cnki.1007-2322.2024.0014

欢迎来到能源电力期刊网！

您当前的位置：

首页 >

文章列表页 >

基于气象辐照数据和引力搜索算法优化VMD-BiLSTM的光伏发电功率预测研究

新能源电力系统 | 更新时间：2026-04-02

- 基于气象辐照数据和引力搜索算法优化VMD-BiLSTM的光伏发电功率预测研究
- Combinatorial Prediction Method Based on Meteorological Irradiation Data and GSA-optimized VMD-BiLSTM
- 现代电力 2026年43卷第2期页码：213-222
- 作者机构：
  
  1. 上海电力大学计算机科学与技术学院, 上海市浦东新区,201306
  2. 国网湖州供电公司,浙江省,湖州市,313000
  3. 上海电力大学电气工程学院, 上海市杨浦区,200090
- 作者简介：
  
  [ "马海洋(2000)，男，硕士研究生，研究方向为电力大数据、新能源发电、人工智能在新型电力系统中的应用，E-mail：yang15421034@163.com" ]
  [ "成贵学(1971)，男，副教授，硕士生导师，通信作者，研究方向为新能源、过程自动化、综合源系统等，E-mail：shiepcheng@shiep.edu.cn" ]
  [ "王函韵(1975)，男，高级工程师，研究方向为电网运行规划与控制，E-mail：wanghanyun@139.com" ]
  [ "赵晋斌(1971)，男，教授，博士生导师，研究方向为现代电力电子技术在电力系统中的应用、新能源发电技术等，E-mail：zhaojinbin@shiep.edu.cn" ]
- 基金信息：
  
  上海市科技创新行动计划项目(21DZ1207502)。
- DOI：10.19725/j.cnki.1007-2322.2024.0014
  中图分类号： 2
- 纸质出版：2026
- 稿件说明：
移动端阅览
马海洋, 成贵学, 王函韵, 等. 基于气象辐照数据和引力搜索算法优化VMD-BiLSTM的光伏发电功率预测研究[J]. 现代电力, 2026,43(2):213-222.

MA Haiyang, CHENG Guixue, WANG Hanyun, et al. Combinatorial Prediction Method Based on Meteorological Irradiation Data and GSA-optimized VMD-BiLSTM[J]. 2026, 43(2): 213-222.
马海洋, 成贵学, 王函韵, 等. 基于气象辐照数据和引力搜索算法优化VMD-BiLSTM的光伏发电功率预测研究[J]. 现代电力, 2026,43(2):213-222. DOI： 10.19725/j.cnki.1007-2322.2024.0014.

MA Haiyang, CHENG Guixue, WANG Hanyun, et al. Combinatorial Prediction Method Based on Meteorological Irradiation Data and GSA-optimized VMD-BiLSTM[J]. 2026, 43(2): 213-222. DOI： 10.19725/j.cnki.1007-2322.2024.0014.

摘要

为提高光伏发电功率预测精度，保证电力系统安全调度和稳定运行，提出一种基于变分模态分解(variational mode decomposition，VMD)和引力搜索算法(gravitational search algorithm，GSA)优化双向长短期记忆神经网络(bi-directional long short-term memory，BiLSTM)的组合预测方法。针对影响光伏发电功率的时序信号存在非线性和非平稳性特征的问题，选取倾斜辐照度、水平辐照度、温度、湿度等气象数据作为特征输入变量，利用VMD算法将其分解为若干个不同的模态并对每个分量分别进行建模；针对传统BiLSTM模型存在参数选择不确定性和收敛速度较慢的问题，引入GSA算法对模型参数优化，缩短手动调制参数的时间，提高超参数设置的精度和效率。使用南方某园区实际光伏发电数据集对所提模型性能进行验证，结果表明，与传统神经网络预测模型和固定参数组合预测模型相比，该方法具有更高的预测准确性和稳定性。

Abstract

To enhance the prediction accuracy of photovoltaic power generation and ensure the safe scheduling and stable operation of the power system

a new combinatorial prediction method is proposed based on variational mode decomposition (VMD) and long short term memory (BiLSTM) optimized by gravitational search algorithm (GSA). To address the issue that the nonlinear and non-stationary characteristics of the temporal signal affect the photovoltaic power generation

meteorological data such as tilt irradiance

horizontal irradiance

temperature and humidity

are selected as the characteristic input variables. These data are subsequently decomposed into several different modes by the VMD algorithm

and each component is modeled individually. To address the uncertainty of parameter selection and slow convergence of traditional BILSTM model

the GSA algorithm is introduced to optimize model parameters. This approach not only shortens the time of manual parameter modulation

but also improves the accuracy and efficiency of hyperparameter configuration. The performance of the proposed model is verified using the actual PV power generation data set of a park in South China. The results indicate that

in comparison to the traditional neural network prediction model and fixed parameter combination prediction model

the proposed method exhibits superior prediction accuracy and stability.

关键词

Keywords

references

MIAO Qingqing, SHI Chunyan, ZHANG Xiangping. Photovoltaic technology under carbon neutrality[J]. Chemical Industry and Engineering Progress, 2022, 41(3): 1125−1131(in Chinese).

DONG Huazhan. Development trend of solar photovoltaic power generation[J]. Modern Industrial Economy and Informationization, 2023, 13(4): 282−284(in Chinese).

WEI Jingdong, ZHAO Zenghai, GUO Yanheng, et al . Status and prospect of China's solar PV generation development in 2021[J ] . Water Power, 2022, 48(10): 4−8(in Chinese).

WANG Chuan. The combination of wind power, photovoltaic power and hydroelectric power[J]. Low Carbon World, 2022, 12(3): 50−52(in Chinese).

DONG Cun, WANG Zheng, BAI Jieyu, et al . Review of ultra-short-term forecasting methods for photovoltaic power generation[J ] . High Voltage Engineering, 2023, 49(7): 2938−2951(in Chinese).

LÜ Qingquan, ZHANG Zhenzhen, MA Yanhong, et al . Analysis and research on output characteristics of regional photovoltaic power generation[J ] . Power Generation Technology, 2022, 43(3): 413−420(in Chinese).

SHU Sheng, XIE Yingming, YANG Wenyu, et al . A review of photovoltaic power generation forecasting methods[J ] . Journal of Engineering For Thermal Energy and Power, 2020, 35(11): 1−11(in Chinese).

ZOU Penghui. Research on method of eliminating fluctuation of photovoltaic generation based on linear auto-disturbance control [J]. Journal of Electric Power, 2020, 35(6): 507−514(in Chinese).

WANG Chenqi, CHEN Xiaodong, CAO Hanwen, et al . Short term photovoltaic power prediction[J ] . Journal of Liaoning Technical University(Natural Science), 2023, 42(1): 99−105(in Chinese).

ZHENG Fang, WANG Huan, CHEN Guoping, et al . Research on the influence and power prediction of distributed photovoltaic access distribution network based on big data[J ] . Light Source and Illumination, 2023(8): 138−140(in Chinese).

LI Baoju, QI Hongwei, FU Jiyue, et al . Analysis on the impact of extreme weather on new energy operation[J ] . Jilin Electric Power, 2022, 50(1): 10−13(in Chinese).

CHANG Yafeng. Application practice of distributed photovoltaic power generation system[J]. Telecom Power Technology, 2022, 39(19): 16−19(in Chinese).

SONG H, AL KHAFAF N, KAMOONA A, et al . Multitasking recurrent neural network for photovoltaic power generation prediction[J ] . Energy Reports, 2023, 9: 369−376.

YANG Shuang, LUO Diansheng, HE Hongying, et al . Output power prediction of photovoltaic power generation system based on EMD-LSSVM model[J ] . Acta Energiae Solaris Sinica, 2016, 37(6): 1387−1395(in Chinese).

ABDELMOULA I A, ELHAMAOUI S, ELALANI O, et al . A photovoltaic power prediction approach enhanced by feature engineering and stacked machine learning model[J ] . Energy Reports, 2022, 8: 1288−1300.

WANG Fuzhong, WANG Shuaifeng, ZHANG Li. Ultra-short-term power prediction of photovoltaic power generation based on VMD-LSTM and error compensation[J]. Acta Energiae Solaris Sinica, 2022, 43(8): 96−103(in Chinese).

LÜ Daqing, YANG Huanhong, DU Haoliang, et al . Health status assessment and prediction of operational condition of a measurement and control device in a UHV converter station based on KPCA and Bi-LSTM[J ] . Power System Protection and Control, 2022, 50(19): 80−87(in Chinese).

CAI Dongyang, ZHAO Shen, ZHOU Wei, et al . Load forecasting based on weather sensitive load decomposition and LSSVM under influence of photovoltaic[J ] . Computing Technology and Automation, 2020, 39(4): 81−85(in Chinese).

BIAN Haihong, SUN Jianshuo. Photovoltaic power generation prediction model based on optimized TMY method-GRNN[J]. Electric Power Engineering Technology, 2021, 40(5): 94−99(in Chinese).

ZUO Siyu, ZHAO Qiang, ZHANG Bing, et al . Prediction of ship motion attitude based on VMD-SSA-GRU model[J ] . Ship Science and Technology, 2022, 44(23): 60−65(in Chinese).

JIANG Zhihang, WANG Yanxia, YAN Jiajun, et al . Modeling and analysis of cotton price forecast based on BILSTM[J ] . Journal of Chinese Agricultural Mechanization, 2021, 42(8): 151−160(in Chinese).

LI Zheng, LI Junhong, GU Juping, et al . Model identification of papermaking process based on gravitational search algorithm[J ] . Control Engineering of China, 2022, 29(11): 1974−1980(in Chinese).

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于变分模态分解的卷积长短时记忆网络短期电力负荷预测方法

基于变分模态和改进多元宇宙优化的短期风电功率预测

基于变分模态分解的卷积神经网络−双向门控循环单元−多元线性回归多频组合短期电力负荷预测

基于时序分解及机器学习的风电功率组合预测模型

基于自适应噪声完备集合经验模态分解–样本熵–双向长短期记忆网络的短期光伏功率预测