【作者】 史建国；

【导师】 张长铠；

【作者基本信息】 山东大学 ， 微生物学， 2006， 博士

【摘要】 我国发酵工业是一个品种繁多，门类齐全，具有相当规模的独立工业体系。产品包括酒类、有机酸、抗生素、酶制剂、淀粉糖、氨基酸、核苷酸、维生素、有机溶剂、微生物杀虫剂、植物激素、单细胞蛋白等。应用领域涉及医药、卫生、轻工、化工、农业、能源、环保等诸多行业；其中，味精、柠檬酸、酶制剂、酵母产量居世界前列。发酵工业已成为国民经济的重要支柱产业之一。 发酵过程的在线检测和自动控制是发酵工业技术进步的重要发展方向。传感器技术作为发酵过程信息的发生源，对实现发酵过程的在线检测和自动控制起到关键作用。然而，目前我国发酵生产上发酵过程自动化控制程度不高，落后于其他领域。用于发酵生产中的传感器主要是进行罐内物化参数的测定。此类传感器的性能较稳定，应用也较为普遍，实现了部分参数的在线监控。但与发酵最优化的自动控制目标相去甚远，即难以成功建立对培养过程进行系统的反馈性控制。因为发酵过程是一个非性线、多变量和随机性的动态过程，发酵体系是一个复杂的被控对象。温度、溶氧、pH、培养基成分、细胞形态、细胞浓度、产物组成及含量等均是发酵过程的重要控制参数。随着计算机及控制技术的突飞猛进，生物传感器技术的发展，发酵动力学模型研究的完善，发酵过程控制系统愈来愈多，应用范围亦越来越广。但是，工业上实现发酵过程最优化自动控制的实例却不多，仍以人工控制和半自动控制为主。其发展滞后的重要原因之一就是基于细胞代谢的生化参数信息的缺乏。 葡萄糖(还原糖)作为微生物的主要碳源和能源，是发酵过程中重要的生化控制参数，它在培养基中的含量及在发酵过程中的浓度变化直接影响发酵产品的质量、收率和生产成本。目前实际生产过程中常采用传统的手工滴定法或比色法进行还原糖的测定，这些方法操作繁琐、人为误差大，给企业生产管理带来很多麻烦，也严重阻碍了发酵过程在线检测和自动控制技术的应用。解决还原糖测定的传感器技术成为发酵工业迫切需要解决的技术难题更多还原

【Abstract】 Fermentation industries, including the production of alcohol, organic acids, enzyme preparations, antibiotics, starch sugars, amino acids, vitamins, bio-pesticides, plant growth regulators, single cell protein etc. have been extensively developed in China and their application covers the field of medicine, food, chemicals, agriculture, energy resource, environmental protection etc.. The scale of some products such as monosodium glutamate(MSG), citric acid, enzyme preparation and yeast occupies a leading position in the world. They are one of the most important industrials in Chinese economy.Fermentation process monitor and control are an important part in bioprocess engineering. Sensors play a key role in acquiring data from a certain bio-process and keeps on updating a control model. The current sensors have been developed to monitor the parameters of temperature, dissolved oxygen, pH, pressure, viscidity, turbidity conveniently and some of which have been applied successfully to online control and optimization systems in fermentation industries, which is assumed to be fully characterized by a constructed model. However, because of multiple variables, serious nonlinear and multiple work modes, many industrial process are too complex to be described by a model only based on physical parameters. Furthermore, the present models not only reduces the estimate precision but also increases the computational complexity, because there are lack of effective online instrument for efficiently measuring the biochemical information that are essential for the fermentation process control.更多还原