人工神经网络在海洋微生物脂肪酶发酵中的应用

Application of artificial neural network in marine microbial lipase fermentation

由于微生物生长过程具有高度的非线性和时变性而且重复性差,使发酵过程控制问题变得很复杂。以海洋酵母Bohaisea-9145发酵生产脂肪酶为例,利用人工神经网络控制方法,建立了批式发酵模型。通过对模型中酶活力、菌体浓度和葡萄糖浓度的模拟和实测值比较,该模型的泛化能力较好,模型对发酵过程的仿真是有效的。以此模型模拟的葡萄糖流加补料策略为方案,脂肪酶发酵活力比批式发酵提高了3.45倍。

Due to the microbial growth process is highly nonlinear and time-varying and poor repeatability, the fermentation process control becomes very complex. In this paper, the marine yeast Bohaisea-9145 producing lipase as the test strain, the batch fermentation feeding strategy was optimized by artificial neutral network and genetic algorithms. By comparing the simulated and measured value of the enzyme activity, cell concentration and glucose concentration, the simulation model of fermentation process is effective. Taking the simulation model of glucose feeding strategy, the lipase activity of feeding fermentation is 3.45 times than the batch fermentation.