[1]吴江,薛金花,余涛,等.基于NARX神经网络-小波分解光伏发电功率预测[J].郑州大学学报(工学版),2020,41(06):79-84.[doi:10.13705/j.issn.1671-6833.2020.06.015]
　Shi Ruxin,Wang Dushun,Yu Tao,et al.Prediction of Photovoltaic Power Generation Based on NARX Neural Network-Wavelet Decomposition[J].Journal of Zhengzhou University (Engineering Science),2020,41(06):79-84.[doi:10.13705/j.issn.1671-6833.2020.06.015]

点击复制

基于NARX神经网络-小波分解光伏发电功率预测()

分享到：

《郑州大学学报(工学版)》[ISSN:1671-6833/CN:41-1339/T]

卷:

41卷

期数:

2020年06期

页码:

79-84

栏目:

出版日期:

2020-12-31

文章信息/Info

Title:

Prediction of Photovoltaic Power Generation Based on NARX Neural Network-Wavelet Decomposition

作者:

吴江; 薛金花; 余涛; 冯鑫振; 王德顺; 窦春霞; 史如新

国网江苏省电力有限公司常州供电分公司,江苏常州213000, 中国电力科学研究院有限公司南京分院,江苏南京210009, 南京邮电大学先进技术研究院,江苏南京210023

Author(s):

Shi Ruxin;  Wang Dushun;  Yu Tao;  etc.

State Grid Jiangsu Electric Power Co., Ltd. Changzhou Power Supply Branch, Jiangsu Changzhou 213000, China Electric Power Science Research Institute Co., Ltd. Nanjing Branch, Jiangsu Nanjing 210009, Advanced Technology Research Institute of Nanjing University of Posts and Telecommunications, Jiangsu Nanjing 210023

关键词:

光伏发电功率预测;  小波分解;  NARX神经网络;  小波重构;  BP神经网络

Keywords:

Power forecast for photovoltaic power generation;  small wave decomposition;  narx neural network;  wavelet reconstruction;  BP neural network

DOI:

10.13705/j.issn.1671-6833.2020.06.015

文献标志码:

A

摘要:

由于光伏发电功率所具有的间歇性、波动性和随机特性，对其进行更加准确的预测可以有效减少光伏发电对电力系统的不利影响，对电力系统的安全经济稳定运行具有重要意义.本文提出了基于小波分解-NARX神经网络组合预测方法，通过小波分解将历史光伏序列分解为高频和低频分量，将高、低频数据作为NARX神经网络输入、光伏输出功率作为神经网络输出进行训练得到预测输出，随后对其进行小波重构推求出光伏发电预测数据.通过仿真结果表明，新的预测算法预测误差比传统BP神经网络更小、预测精度更高，并且具有良好的适应性，并证实了基于小波分解-NARX神经网络组合预测方法的可行性和高效性.

Abstract:

Due to the intermittent, volatility and random characteristics of photovoltaic power generation, more accurate prediction can effectively reduce the adverse impact of photovoltaic power generation on the power system, which is of great significance to the safe and economical operation of the power system. Based on the wavelet decomposition-NARX neural network combined prediction method, the historical PV sequence is decomposed into high frequency and low frequency components by wavelet decomposition, and the high and low frequency data are used as the NARX neural network input and the PV output power is used as the neural network output to train and obtain the predicted output. Then the wavelet reconstruction is used to derive the PV power generation prediction data. The simulation results show that the prediction algorithm of the new prediction algorithm is smaller than the traditional BP neural network, the prediction accuracy is higher, and it has good adaptability. The feasibility and efficiency of wavelet decomposition-NARX neural network combined prediction method

常用功能