摘要
根据模拟移动床(SMB)的发展现状和操作流程要求,对模拟移动床色谱分离的工艺流程进行了重新设计,使得系统的整个操作流程更为简单实用。用BP神经网络对苯丙氨酸模拟移动床的色谱分离过程进行了建模,用实数编码的遗传算法对分离参数和操作条件进行了优化,并在实际生产中得到应用,使产品的纯度和回收率都得到了较大的提高。该工艺具有分离效果好等优点,现已在江阴某公司成功投入使用,苯丙氨酸的纯度和回收率都达到了98%以上。
Based on the study of developing actuality and operation flow of SMB （simulated moving bed）, a technical flow of SMB chromatngram separation is designed renewedly, which makes the operating flow simpleness and practicality. With BP neural network and genetic algorithm, the model of phenylalanine SMB separation is established, the separate parameter and operating condition are optimized, it advances the purity and recovery in practice. This technics has the excellence of good separation result, etc. it was applied successfully in Jiangyin, the purity and recovery of phenylalanine exceed 98%.
出处
《计算机测量与控制》
CSCD
2005年第8期796-798,823,共4页
Computer Measurement &Control
关键词
模拟移动床
工艺流程
神经网络
遗传算法
simulated moving bed
technical flow
neural network
genetic algorithm
