A NEURAL ESTIMATOR FOR AN ERYTHROMYCIN FED-BATCH FERMENTATION PROCESS

摘要：

In this paper, a neural estimator is designed for optimal control of an erythromycin fed-batch fermentation process. To achieve good product quality, the hypha, total sugar and erythromycin densities have to be estimated accurately. However, the three variables can not be measured directly. Moreover, the whole process exhibits sever nonlinearity and is time-varying and subject to large unknown disturbances. To tackle these problems, a neural network is used to estimate the three variables based on the measured pH and dissolved oxygen levels in the fermentation process. The experimental result shows that the trained neural network have achieved satisfactory estimation under different testing conditions, therefore can be used for realtime measurement of the densities of hypha, total sugar and end product in the process.

关键词： Erythromycin fermentation Neural Networks Estimation.

作者: Y.J. Di M.R. Fei Z.Z. Chen

作者单位: Shanghai Key Laboratory of Power Station Automation Technology,200072 Shanghai,China

会议类型: 国际会议

会议名称: 2008高等智能国际会议(2008 International Conference on Advanced Intelligence)

会议地点: 北京

会议语种:英文

在线出版日期: 2008-10-18（万方平台首次上网日期，不代表论文的发表时间）