基于RBF神经网络的青霉素发酵过程的模型辨识

Model building in the penicillin fermentation process based on RBF neural networks for identification

分析了目前青霉素发酵过程中存在的问题.基于RBF神经网络的辨识方法,建立了青霉素发酵过程模型.以动力学模型为基础产生教师数据,采用遗传算法对网络进行训练,建立了基于RBF神经网络的发酵过程模型,并进行了仿真实验验证.实验结果表明,该辨识模型对青霉素补料分批培养过程具有实用价值.

A new way of model building for identification based on RBF neural networks in the penicillin fermentation process was presented .The penicillin fermentation process is a process of time varying and nonlinear feature .RBF neural networks have the ability to get a nonlinear function. Genetic algorithms were used to train the neural networks .A simulation was made to prove this fermentation model . The result of the simulation shows that the model is practicable.