人工神经网络在棒束临界热流密度预测中的应用  

Application of Artificial Neural Network in Bundle Critical Heat Flux Prediction

作　　者：韩浪[1] 冉旭[1] 单建强[1] 黄彦平[2] 陈炳德[2] 朱继洲[1] 

机构地区：[1]西安交通大学核能与热能工程系,陕西西安710049 [2]中国核动力研究设计院,四川成都610041

出　　处：《原子能科学技术》2006年第3期257-261,共5页Atomic Energy Science and Technology

摘　　要：基于已有的棒束临界热流密度数据库,采用COBRA-Ⅳ程序计算得到子通道局部临界热流密度数据库。用人工神经网络(ANN)理论对数据库进行训练,得到基于ANN理论的棒束临界热流密度预测模型。预测模型的预测精度显著高于常用经验关系式的预测精度,其预测值的均方差为5.63%。A bundle critical heat flux（CHF） database based on subchannel local condition is obtained by analyzing existing bundle experimental database with COBRA IV code. Artificial neural network is then applied to train the database and a bundle CHF prediction model is finally obtained. The prediction accuracy of the obtained model is much better than that from general empiric formula, and the root-mean-square of predicated value is 5.63 %.

关 键 词：临界热流密度 人工神经网络 棒束 

分 类 号：TL333[核科学技术—核技术及应用]