摘要
本文提出了一种发酵过程混合建模方法。该模型由基于质量平衡的发酵过程机理方程与基于数据的支持向量机模型2部分组成。机理方程作为描述过程的动态行为的整体框架,支持向量机模型用来进行模型参数的估计。并且提出了一种在线修正策略,增强了模型的实用性。将所建立的混合模型应用于诺西肽发酵过程中,结果表明,混合建模方法具有很好的估计性能。
A hybrid modeling scheme was developed for fermentation process, which combines the first principle equation based on mass balance and data based support vector machine model. First principle describes the rough framework of fermentation process. And support vector machine model is used to estimate the unmeasured process parameters that are difficult to model with first principle. Also, an on-line correction strategy was proposed to improve the model's practicality. Applying the proposed modeling scheme to Nosiheptide fermentation process demonstrates that the hybrid modeling has good estimation performance.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2007年第1期34-37,共4页
Chinese Journal of Scientific Instrument
基金
国家自然科学基金(60374003)
973计划子课题(2002CB312200)资助项目