Bio-fermentation System Algorithm Based On Intelligent Control Research And Practice

With the rapid development of the economy, there are more and more systems with nonlinear time varying and long time lag being used in the industrial process, the traditional control methods no longer satisfy those complex systems'requirements. The intelligence control systems are needed more urgently than ever before. With the advantages of the micro-electronics and the computer technology, the intelligence control theory can be much conveniently used in industrial process. However, most of the related researches are focused on simulation research or theory research and their successful applications in industry are rarely mentioned.The paper takes the fermentation process of producing human-like collagen by recombinant E. coli as its controlled object, a set of intelligence control system was designed and constructed for increasing the products rate and improving the product's quality, where the BP neural network algorithm was introduced through utilizing computer technology, combining the control system hardware with computation algorithm. Since there are no reasonable theory in deciding hide nodes in BP neural network theory, an improved experimental formula was created and the influence of the number of sample data on how many hide nodes needed was concerned, the inadequacy of original experimental formula was compensated. The experiment results showed that the characteristics of the system are quite satisfactory and the expected requirements are obtained.The paper takes the combination of the control science with the microbiology as an entry point, by using the experimental platform mentioned above, puts forward a new control strategy which is quite pioneering. On the other hand, from the view of application of the theory, this thesis gives a useful example of the intelligence theory's application for solving control of complex systems, and extends its application field.