基于注意力机制的Encoder-Decoder光伏发电预测模型  

Encoder-Decoder Photovoltaic Power Generation Prediction Model Based on Attention Mechanism

作　　者：宋良才 索贵龙 胡军涛 窦艳梅 崔志永 SONG Liang-cai;SUO Gui-long;HU Jun-tao;DOU Yan-mei;CUI Zhi-yong(Company of Henan Hebi National Grid Power Supply, Hebi 458000, China;School of Computer, Wuhan University, Wuhan 430000, China)

机构地区：[1]国网河南省电力公司鹤壁供电公司,河南鹤壁458000 [2]武汉大学计算机学院,湖北武汉430000

出　　处：《计算机与现代化》2020年第9期112-117,共6页Computer and Modernization

基　　金：国网河南省电力公司科技项目(SGHAHB00XTJS1900147)。

摘　　要：影响光伏发电系统出力的天气因素具有很大的波动性和不连续性,因此需要创建合适的预测模型来对光伏出力特性进行精准预测,从而保证电网系统的有效运行。本文通过最大信息系数选择合适的历史光伏发电数据,将其作为特征之一进行输入数据重构,并在由LSTM神经元构建的Encoder-Decoder模型上引入注意力机制,最终得到结合注意力机制的Encoder-Decoder光伏发电预测模型。经实际光伏电厂算例分析,验证了所提模型在光伏发电预测方面的准确性和适用性。The weather factors that affect the output of photovoltaic power generation systems have great volatility and discontinuities.Therefore,it is necessary to create a suitable prediction model to accurately predict the characteristics of photovoltaic output to ensure the efficient operation of the power generation network.This paper selects the appropriate historical photovoltaic power generation data through the maximal information coefficient,uses it as one of the features to reconstruct the input data,and attention mechanism is introduced on the Encoder-Decoder model constructed by LSTM neurons to obtain an attention-based Encoder-Decoder photovoltaic power generation prediction model.The analysis of actual photovoltaic power plant examples verifies the accuracy and applicability of the proposed model in predicting photovoltaic power generation.

关 键 词：光伏发电 最大信息系数 长短期记忆神经网络 Encoder-Decoder框架 注意力机制 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]