Parameter Identification For A Class Of Microbial Fed-batch Fermentation Process

In recent years,scholars all over the world have attached great importance to the complicated biological process of glycerol fed-batch fermentation to produce 1,3-PD,and carried out a series of researches,and also obtained certain research results.With the development of the times and the improvement of the scientific research level,the accuracy requirements of the corresponding parameters in the model are getting higher and higher,which requires us to select a better method to identify the parameters according to the characteristics of the model,so that the calculated values of the model are closer to the experiment data.The main arrangements and results of this paper are as follows:1.A parameter optimization dynamic optimization model was constructed for the parameter identification of microbial fed-batch fermentation process.In order to effectively solve the dynamic optimization model,the differential equations in the dynamic optimization model are first transformed into algebraic equations based on the modified configuration formula and the Runge-Kutta formula,respectively,so that the original dynamic optimization problem is transformed into a general nonlinear programming problem,and then applied in nonlinear programming.The effective set algorithm solves the transformed nonlinear programming problem,obtains the optimal parameter values of the parameter identification problem,and analyzes and compares the calculation results.Compared with the existing identification results in the literature,the method has obtained a smaller error.2.In this paper,a two-stage method is proposed to solve the parameter identification problem of the above-mentioned microbial flow fermentation process.First,in the first stage,the modified configuration formula is applied to transform the differential equations in the model into algebraic equations,and then the transformed algebraic equations are rewritten as relatively simple forms.Therefore,the original parameter identification dynamic optimization problem is transformed into a general nonlinear programming problem and solved;Then solve a nonlinear equation in the second stage,and get the optimal parameters of the original parameter identification problem.The calculation results show that the two-stage method can better identify the parameters.