The Research On Modeling And Feeding Optimization Of The Fermentation Process Based On Dividing Stages

The fermentation process is a very complex process of biochemical reaction, which has the characteristics of nonlinear, time-varying and uncertainty. Pichia fermentation occupies an important position in the fermentation industry, which can produce a lot of recombinant protein products whose structure is complicated and have difficulty to synthesize. Pichia expression of recombinant protein fermentation process acts as the research background, fermentation process is established to model in stages and optimize the rate of feeding.First of all, through referring to related literature of the fermentation process modeling in stage and feeding rate optimization, fermentation process has the problem that key parameters are difficult to detect and determining feeding rate by artificial experience is inaccurate. By analyzing the advantages and disadvantages of the algorithm, artificial neural network is used to model and genetic algorithm is adopted to optimize the rate of feeding. Then the basic principle and the operation process of the neural network and genetic algorithm are introduced, and the idea of design and operation process of the combining artificial neural network and genetic algorithm are highly analysed.Secondly, the fermentation process of the pichia expressing recombinant protein is detailly analysed and the influence of various parameters on the fermentation process is studied. According to the basic characteristics of fermentation process, the fermentation process is divided into two stages, bacteria growth stage and product generation stage. Based on the characteristics of each stage, BP neural network is used to respectively establish the models of bacteria growth stage and product generation stage.Finally, the fermentation way is systematically analysed and the fermentation way of fed-batch about pichia expressing recombinant protein is researched. Based on the model established by neural network, the genetic algorithm is adopted to respectively optimizate ammonia feeding rate and glycerol feeding rate of the bacteria growth stage and the ammonia feeding rate and methanol feeding rate of the product generation stage.