摘要
在 150mm流化床上,研究了煤与典型组分垃圾流化床焚烧Cl→HCl转化率与床温和氧量的关系.参照试验和文献数据,采用典型相关理论详细阐述了BP神经网络在垃圾流化床焚烧HCl排放预测的建模过程.模拟表明,模型最佳网络拓扑结构为14×6×1.实际预测表明模型预测精度高、泛化能力好.
Cl→HCl conversion level of typical wastes varies with bed temperature and oxygen was studied in the 150 mm fluidized bed.This paper demonstrates firstly the use of artificial neural networks to establish HCl prediction model based on other references and canonical correlation analysis.The optimal topological structure of model is 14×6×1. The simulated results indicate that model have high accuracy and good generalization ability.
出处
《煤炭学报》
EI
CAS
CSCD
北大核心
2004年第3期333-337,共5页
Journal of China Coal Society
基金
国家自然科学基金资助项目(N59836210)
关键词
流化床
HCL
BP神经网络
典型相关分析
fluidized bed
HCl
artificial neural networks
canonical correlation analysis
