摘要
针对风电功率数据的时序性特点,提出降噪循环神经网络模型对电场中短期内的风电功率进行预测.通过模型能够挖掘其蕴含的知识,提高电力系统的稳定性,优化电力调度.模型首先采用循环神经网络构建一个编码-解码结构,设计编码器从序列变量中获取相应的深度特征,再通过解码器对深度特征进行解码,还原输入序列的状态并预测下一时刻的输出.进而,模型在解码器中设计降噪模块和预测模块,克服一般循环神经网络难以对带噪声数据进行预测的问题,使得模型能够对含有噪声的输入变量进行分析.通过利用电力物联网所采集的数据进行实验,结果证明提出的方法能很好地对风电功率进行预测,达到较好的预测效果.
Aiming at time series characteristics of wind power data,the Denoising Recurrent Neural Network model is proposed to predict the wind power in the short and medium term of the electric field.Through the model,the knowledge contained in it can be mined to improve the stability of the power system and optimize the power dispatching.An encoding-decoding structure is first designed in the recurrent neural network model,and the encoder is designed to obtain corresponding depth features from the sequence variables,and then the decoder decodes the depth features,restore the state of the input sequence and makes a prediction.Furthermore,the model designs the denoising module and prediction module in the decoder to overcome the difficulty of predicting noisy data with traditional recurrent neural networks,enabling the model to analyze input variables containing noise.By using the data collected by the power Internet of things to conduct experiments,the results show that the proposed method can forecast the wind power well and achieve a better prediction effect.
作者
田增垚
彭飞
孟庆东
王汉军
田长翼
陈志奎
TIAN Zeng-yao;PENG Fei;MENG Qing-dong;WANG Han-jun;TIAN Chang-yi;CHEN Zhi-kui(Northeast Branch of State Grid Corporation of China,Shenyang 110180,China;Shenyang Institute of Computing Technology Co.Ltd,Chinese Academy of Sciences,Shenyang 110168,China;School of Software Technology,Dalian University of Technology,Dalian 116620,China)
出处
《微电子学与计算机》
2021年第3期27-32,共6页
Microelectronics & Computer
基金
国家自然科学基金面上项目(61672123)
国家电网公司项目(26992618008K)。
关键词
风电功率预测
时间序列
序列模型
编码-解码结构
降噪循环神经网络
wind power prediction
time series
sequence model
encoder-decoder structure
denoising recurrent neural network
