基于神经网络的油田光伏发电预测方法
Prediction Method of Oilfield Photovoltaic Power Generation Based on Neural Network
郑广围1
作者信息
- 1. 大庆油田装备制造集团石油专用设备有限公司,黑龙江 大庆 163000
- 折叠
摘要
针对油田大规模发展光伏发电时,电网输出功率不稳定的问题,提出了基于神经网络的油田光伏发电预测方法.首先,对光伏发电数据进行归一化和反归一化的预处理;其次,构建带有输入层、隐含层以及输出层的BP神经网络预测模型;最后,根据光伏发电数据的对比序列和参考序列得出每个时刻的光伏数据权重,根据权重提取数据并代入预测模型,实现油田光伏发电预测.结果表明,所提方法的预测精度较高.
Abstract
Aiming at the problem of unstable output power of the power grid during the large-scale development of photovoltaic power generation in oil fields,a neural network-based method for predicting photovoltaic power generation in oil fields is proposed.Firstly,the photovoltaic power generation data is preprocessed by normalization and de-normalization.Secondly,a BP neural network prediction model with input layer,hidden layer,and output layer is constructed.Finally,the photovoltaic data weights at each time point are obtained based on the comparison sequence and reference sequence of the photovoltaic power generation data.Based on the weights,the data is extracted and substituted into the prediction model to achieve oilfield photovoltaic power generation prediction.The results show that the proposed method has high prediction accuracy.
关键词
神经网络/光伏发电/油田光伏Key words
neural network/photovoltaic power generation/oilfield photovoltaic引用本文复制引用
出版年
2024
NETL
NSTL
万方数据