摘要
结合某电厂垂直水冷壁壁温频繁超温问题,采用垂直水冷壁壁温超温机理和门控神经网络算法(GRU)相融合的方法,建立壁温预测模型,采集电厂历史运行数据,经数据预处理生成数据集,统计与分析预测模型数据训练结果,确定研究与优化方向,提高预测模型的精度和鲁棒性。
Combined with the frequent overtemperature problem of the vertical water wall in power plant,the overtemperature mechanism of vertical water wall and the gated neural network algorithm(GRU) are integrated to establish the prediction model,collect the historical operation data of power plant and perform data preprocessing,based on the statistics and analysis of data training results,develop further research and optimization direction to improve the accuracy and robustness of prediction model.
作者
吴戈杨
魏国华
凌朝年
路丕思
WU Geyang;WEI Guohua;LING Chaonian;LU Pisi(Guangdong Datang International Leizhou Power Generation Co.,Ltd.,leizhou 524255,China;State Key Laboratory of Low-carbon Thermal Power Generation Technology and Equipments(Harbin Boiler Company Limited),Harbin 150046,China)
出处
《锅炉制造》
2023年第5期4-6,12,共4页
Boiler Manufacturing
关键词
垂直水冷壁壁温超温
预测模型
数据挖掘
GRU
the overtemperature problem of vertical water wall
prediction model
data mining
GRU
