引用本文: | 杨强大,王福利,常玉清.基于阶段辨识的诺西肽发酵过程菌体浓度软测量[J].控制理论与应用,2009,26(9):1026~1030.[点击复制] |
YANG Qiang-da,WANG Fu-li,CHANG Yu-qing.Phase-identifying-oriented soft sensor for biomass in Nosiheptide fermentation process[J].Control Theory and Technology,2009,26(9):1026~1030.[点击复制] |
|
基于阶段辨识的诺西肽发酵过程菌体浓度软测量 |
Phase-identifying-oriented soft sensor for biomass in Nosiheptide fermentation process |
摘要点击 2117 全文点击 1357 投稿时间:2007-09-26 修订日期:2009-04-27 |
查看全文 查看/发表评论 下载PDF阅读器 |
DOI编号 10.7641/j.issn.1000-8152.2009.9.018 |
2009,26(9):1026-1030 |
中文关键词 发酵 软测量 辅助变量选择 阶段辨识 神经网络 |
英文关键词 fermentation soft sensor selection of secondary variables phase identification neural networks |
基金项目 国家自然科学基金资助项目(50974146); 国家973计划子课题资助项目(2002CB312201). |
|
中文摘要 |
由于发酵过程中系统非线性特性与发酵阶段密切相关的实际特点, 针对诺西肽发酵过程菌体浓度的估计问题, 提出了一种基于阶段辨识的软测量方法. 首先以分阶段的诺西肽发酵过程非结构模型为基础, 根据隐函数存在定理进行辅助变量的合理选择; 然后利用经数学推导得到的指示变量”伪比生长率”完成发酵阶段的在线辨识, 并采用神经网络构建出对应于各阶段的局部软测量模型. 实际应用结果表明, 所提方法有效、预估精度较高. |
英文摘要 |
Fermentation processes have different nonlinear characteristics in different fermentation phases. A Phaseidentifying-oriented soft sensor approach is proposed for estimating the biomass in Nosiheptide fermentation process. Based on the segmented unstructured model of Nosiheptide fermentation process, the secondary variables are selected according to the implicit function existence theorem. The on-line identification of fermentation phases is accomplished by using an indicator variable which is gained by mathematical inference, and for each phase, a local soft sensor model is developed. The practical application results show the effectiveness and validity of the presented approach. |