| 人工神经网络模拟反应精馏过程的研究 |
| |
| 引用本文： | 冯国红,宋海华.人工神经网络模拟反应精馏过程的研究[J].太原科技大学学报,2010,31(1):76-80. |
| |
| 作者姓名： | 冯国红 宋海华 |
| |
| 作者单位： | 1. 太原科技大学材料学院,太原,030024
2. 天津大学化工学院,天津,300072 |
| |
| 摘 要： | 提出了用BP(误差反向传播)神经网络模拟计算合成乙酸甲酯的新思路,模拟过程中采用学习速率可变的动量BP算法训练神经网络。结果表明:只要有充足可靠的数据为基础,采用学习速率可变的动量BP算法训练的神经网络的预测精度比普通BP算法的预测精度高10倍左右,且训练时间显著下降,是一种具有广泛应用前景的模拟方法。
|
| 关 键 词： | 反应精馏 学习速率可变 BP神经网络 模拟 乙酸甲酯 |
Application Research on the Artificial Neural Network in Simulating Reactive Distillation |
| |
| FENG Guo-hong,SONG Hai-hua.Application Research on the Artificial Neural Network in Simulating Reactive Distillation[J].Journal of Taiyuan University of Science and Technology,2010,31(1):76-80. |
| |
| Authors: | FENG Guo-hong SONG Hai-hua |
| |
| Institution: | FENG Guo-hong,SONG Hai-hua(1.Institute of Materials,Taiyuan University of Science Technology,Taiyuan 030024,China,2.School of Chemical Engineering , Technology,Tianjin University,Tianjin 300072,China) |
| |
| Abstract: | It is an innovation to apply ANN(artificial neural network)which trained by momentum BP(back propagation)with variable learning rate to simulate and analysis reactive distillation for synthesizing methyl acetate.The results indicate that if there are enough data,the precision of predicted data from ANN trained by momentum BP with variable learning rate is as 10 times as common and the trained time decreases rapidly.It is a simulative method with wide prospect. |
| |
| Keywords: | reactive distillation variable learning rate BP artificial neural network simulation methyl acetate |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
