Parameter Identification In Microbial Continuous Fermentation Based On The Different Trans-membrane Mode

Taking the microbial continuous fermentation as the background, based on thecharacteristics and metabolic reaction pathways in the continuous fermentationprocess, in order to discuss the trans-membrane transport mode of glycerin,1,3-PDand the effect of3-HPA on enzymes, the dynamic models with intracellular substrateand products are built firstly. Several properties, the existence and uniqueness of thesolution and its continuous dependence on parameters about the systems are provedsecondly. Lastly the relevant parameter identification model is built, and it is solvedby improved particle swarm optimization algorithm to get the best parameters. Themain work and achievements are as follows:Firstly, supposing middle product3-HPA’s effect on microbial growth does notexist, and its effect on two critical enzymes always exists, Based on the existingmicrobial continuous fermentation system, discuss trans-membrane transport modesabout glycerol and1,3-propanediol in nine cases, and respectively improvecorresponding dynamic systems with enzymatic reaction are built in9cases. Discussthe properties of the nonlinear systems, and the existence of the solution, and itscontinuous dependence and the differentiability on the parameters are proved. Takingthe average relative error between the calculated value and the experimental value asthe objective function, taking the dynamic system as the state constraint, build therelevant parameter identification models, and solve these models through improvedparticle swarm optimization algorithm. Through the numerical solution obtained, theconclusion below is given: glycerol transports across the cell membrane by passiveway, and1,3-propanediol transports across the cell membrane by the active andpassive combined way.Secondly, based on the above conclusion: glycerol transports across the cellmembrane by passive way, and1,3-propanediol transports across the cell membraneby the active and passive combined way, consider the effect of3-HPA on two criticalenzymes, build new models and solve the parameters by improved particle swarmoptimization algorithm. The calculation results show that: the effect of3-HPA on twoenzymes is not obvious to the whole model in the current technical conditions.