摘要
以芳烃吸附Parex过程为例,在建立其稳态机理模型的基础上,采用正文设计方法获得模型的特性数据;根据这些数据建立具有一个隐层的神经网络模型,并以此作为优化操作的简化模型。采用一种新的优化方法(SAGACIA),得到吸附塔的最佳操作条件,以实现计算机的二级优化控制。
Taking the adsorption column of a petrochemical plant for the case study,based on characteristic data which is obtained from the mechanism steady-state model by means of orthonormal design method,the artificial neural networks model is investigated.The operation condition of this column is optimized by a new optimization method SAGACIA,The result is satisfactory,and the enterprise will gain more economical benefit.
出处
《华东理工大学学报(自然科学版)》
CAS
CSCD
1996年第6期749-753,共5页
Journal of East China University of Science and Technology
基金
国家自然科学基金
关键词
神经网络
简化模型
操作
优化
移动床
Parex过程
steady-state model
artificial neural networks model
optimization
SAGACIA optimization method