摘要
采用神经网络的方法建立水泥预分解窑煅烧工段的预测模型。选择合理的状态与控制变量,通过采集实际运行数据来训练神经网络。构建的基于BPNN神经网络的煅烧预测模型能够较好地拟合采样数据,具有较好的泛化能力。
Making use of neural network method to establish predictive model for kiln calcination process was discussed,and reasonable state and control variables and collecting actual operation data were chosen to train neural network weights.The established calcination process predictive model based on BP neural network can fit the sampled data better and has good generalization ability.
出处
《化工自动化及仪表》
CAS
2012年第2期156-158,共3页
Control and Instruments in Chemical Industry
基金
安徽省自然科学基金资助项目(10040606Q64)
宿州学院智能信息处理实验室开放课题资助项目(2011YK1-11)
关键词
BPNN
煅烧工段
模型
拟合
泛化
back propagation neural network,calcination process,model,fitting,generalization
