| The penicillin is the biggest demand antibiotic of various countries, the ferment becomes more and more important in the national economy status. The researchers try their best to ameliorate the technology and the equipment of the fermentation, simultaneously they pay more attention on the inspection and control of the fermentative process. But because the penicillin is the secondary metabolite of the microorganism, the mechanism of the fermentative process is extremely complex, and it is a multi-variable, uncertainty, strong coupling and non-linear system, so it is difficult to establish the precise mathematical model, which makes the optimization control of the entire fermentation process becomes very difficult.To solve the problem, we take the penicillin fermentative process as the study subject, and carry a thorough research on the control question of the fermentative process.(1) By unifying the inverse system approach and the neural network, we put forward a fermentative process decoupling control method which is based on the neural network inverse system. According to the characteristic of the penicillin fermentative process, we design the corresponding mathematical model and prove the inverse character of the fermentative system. On this basis we get the pseudo-linear system by structuring the neural network counter system further and bunching the fermentation system. At last we implement the high performance decoupling control by designing the linear closed loop regulator. The simulation result indicates that the decoupling control method based on the inverse system of neural network has a perfect control on the penicillin fermentative process.(2) In this thesis we design a set of complete biological fermentation numerical control system which implement the real-time control of the main parameter during the penicillin fermentative process, and the making-up material control mode based on the neural network counter method.
|
PDF Full Text Request