摘要
针对工业生产过程的复杂性和时变性 ,提出一种用于工业生产过程建模的自适应监督式分布神经网络 (SDNN)。介绍了 SDNN网络的结构和自适应学习方法 ,并将 SDNN网络与传统建模方法相结合 ,应用于铅锌烧结过程的烧结块成分预测。工业应用结果表明 ,SDNN模型具有较高的预测精度 。
Considering the complexity and the time-variability of industrial process, an adaptive Supervised Distributed Neural Networks (SDNN) is proposed for modeling of industrial process. The structure and the adaptive learning methods of the SDNN are introduced. The SDNN combined with conventional modeling methods is applied to prediction of agglomerate's composition in Pb-Zn sintering process. Results of industrial applications show that the SDNN has higher precision, which combined with other modeling methods is effective in modeling of industrial process.
出处
《控制与决策》
EI
CSCD
北大核心
2001年第5期549-552,556,共5页
Control and Decision
基金
国家 8 63 /CIMS项目 (863-511-984 5-0 0 3
863-511-94 5-0 14 )
关键词
监督式分布神经网络
自适应学习
铅锌烧结过程
人工神经网络
智能化
Algorithms
Computer simulation
Forecasting
Lead
Learning systems
Mathematical models
Process engineering
Sintering
Zinc
