基于双向LSTM-Attention模型的火电厂负荷预测研究  

Research on load prediction of thermal power plants based on BiLSTM-Attention model

作　　者：陈恩帅 茅大钧 陈思勤[2] 魏立志 CHEN Enshuai;MAO Dajun;CHEN Siqin;WEI Lizhi(School of Automation Engineering,Shanghai University of Electric Power,Shanghai 200090,China;Huaneng International Power Co.,Ltd.,Shanghai Shidongkou Second Power Plant,Shanghai 200942,China;Siping No.1 Thermal Power Company,Jilin Electric Power Co.,Ltd.,Siping 136001,Jilin China)

机构地区：[1]上海电力大学自动化工程学院,上海200090 [2]华能国际电力股份有限公司上海石洞口第二电厂,上海200942 [3]吉林电力股份有限公司四平第一热电公司,吉林四平136001

出　　处：《电力科技与环保》2024年第4期380-387,共8页Electric Power Technology and Environmental Protection

基　　金：国家自然科学基金项目(62373241);中国华能集团有限公司2022年度科技项目(HNKJ22-HF22)。

摘　　要：准确预测电厂负荷可指导火电厂制定发电计划和调度安排,有利于降低能源成本和污染物排放,对电厂的经济性和环保性有重要意义。本文提出一种基于双向LSTM-Attention的火电厂负荷预测方法。首先,通过皮尔逊系数筛选出关键特征变量;其次利用双向长短期记忆网络提取关键变量之间的长期依赖关系与短期变化特征,最后融合注意力权重机制以进一步突出关键时序信息,进而实现负荷的准确预测。以某在役600 MW超临界机组为对象进行验证。结果表明:相较于单向LSTM、双向LSTM、单向LSTM-Attention,本文所提方法的决定系数R^(2)、均方根误差S_(RMSE)和平均绝对误差S_(MAE)均为最优,分别为0.9566、16.3159、13.5043,能更准确地捕捉到负荷快速波动的趋势,为电厂的负荷预测和能源管理提供可行的方法。Accurate prediction of load can guide thermal power plants to formulate power generation plans and scheduling arrangements,which is conducive to their reduction of energy costs and pollution emissions,and is of great significance to the economy and environmental protection of power plants.Therefore,a load forecasting method of thermal power plant based on BiLSTM-Attention is proposed in this paper.Firstly,the key characteristic variables are screened by Pearson coefficient.Secondly,BiLSTM was used to extract the long-term dependence relationship and short-term change characteristics among key variables,and finally,the Attention mechanism was integrated to further highlight the key timing information,so as to achieve accurate load prediction.A 600 MW supercritical unit in service was used for validation.Compared to LSTM,BiLSTM,LSTM-Attention,the results of BiLSTM-Attention show that the coefficient of determination R^(2),root mean square error S_(RMSE)and mean absolute error S_(MAE)are optimal,they are 0.9566、16.3159、13.5043,which can more accurately capture the trend of rapid load fluctuation,it can take the accurate prediction of load for thermal power plants.

关 键 词：火电厂 负荷预测 双向LSTM模型 Attention机制 能源管理 

分 类 号：TK011[动力工程及工程热物理]