辽宁科技大学学报
遼寧科技大學學報
료녕과기대학학보
Journal of University of Science and Technology Liaoning
2015年
4期
315-320
,共6页
金光迪%李晓红%王艳%冯恩民%王勇%吕荣建
金光迪%李曉紅%王豔%馮恩民%王勇%呂榮建
금광적%리효홍%왕염%풍은민%왕용%려영건
批式流加发酵%非线性动力系统%参数辨识
批式流加髮酵%非線性動力繫統%參數辨識
비식류가발효%비선성동력계통%삼수변식
fed-batch fermentation%nonlinear dynamical system%parameter identification
微生物批式流加发酵生产1,3-丙二醇是一种重要的生产方式,但是目前研究只考虑了细胞外5种物质浓度的变化,没有考虑代谢机理、细胞内物质浓度的变化及重要的中间产物.本文将细胞内甘油、3-羟基丙醛和1,3-丙二醇三种物质的浓度变化考虑在内,假设甘油跨膜运输情形为主动运输和被动扩散,1,3-丙二醇的跨膜运输情形为主动运输和被动扩散,建立了微生物批式流加发酵动力系统.以实验值与计算值的误差平方和作为性能指标建立参数辨识模型,并用改进的粒子群算法求解模型中参数.结果表明,改进后模型平均相对误差比现有文献略有降低,说明了模型的有效性.
微生物批式流加髮酵生產1,3-丙二醇是一種重要的生產方式,但是目前研究隻攷慮瞭細胞外5種物質濃度的變化,沒有攷慮代謝機理、細胞內物質濃度的變化及重要的中間產物.本文將細胞內甘油、3-羥基丙醛和1,3-丙二醇三種物質的濃度變化攷慮在內,假設甘油跨膜運輸情形為主動運輸和被動擴散,1,3-丙二醇的跨膜運輸情形為主動運輸和被動擴散,建立瞭微生物批式流加髮酵動力繫統.以實驗值與計算值的誤差平方和作為性能指標建立參數辨識模型,併用改進的粒子群算法求解模型中參數.結果錶明,改進後模型平均相對誤差比現有文獻略有降低,說明瞭模型的有效性.
미생물비식류가발효생산1,3-병이순시일충중요적생산방식,단시목전연구지고필료세포외5충물질농도적변화,몰유고필대사궤리、세포내물질농도적변화급중요적중간산물.본문장세포내감유、3-간기병철화1,3-병이순삼충물질적농도변화고필재내,가설감유과막운수정형위주동운수화피동확산,1,3-병이순적과막운수정형위주동운수화피동확산,건립료미생물비식류가발효동력계통.이실험치여계산치적오차평방화작위성능지표건립삼수변식모형,병용개진적입자군산법구해모형중삼수.결과표명,개진후모형평균상대오차비현유문헌략유강저,설명료모형적유효성.
The method of microbial fermentation is an important mode of production in producing 1,3-Propane-diol (1,3-PD). Although the microbial fed-batch culture of glycerol to 1,3-PD is a vital production mode, the present research results of microbial fed-batch culture system which only think about the concentration of five kinds extracellular materials changed instead of metabolic mechanism, concentration of intracellular materials changed and important intermediate products. We take the concentration of intracellular glycerol, 3-HPA and 1, 3-PD into consideration in this paper. A dynamic system of microbial fed-batch fermentation is built under the active transport and passive diffusion of glycerol and active transport and passive diffusion of 1,3-PD. Taking the sum of square error between experimental data and computational values as the performance index, the pa-rameter identification model is presented and an improved particle swarm optimization algorithm is developed to seek the unknown parameters in the model. Numerical simulations show that the average relative errors of biomass and glycerol have reduced slightly and illustrate the availability in the improved model.