摘要
提出一种简易的真菌深层培养过程网络模型。输入变量为可在线测量的排气中的二氧化碳浓度,网络权数采用遗传算法进行优化训练。所获神经网络模型能准确预测培养过程的状态变量(生物量浓度,产物浓度等)。研究表明遗传算法训练此类神经网络系统是可行的。
A Simple model based on Neural Nework is proposed for optimizing the process of fungus suspended culture. Concentrations of CO2 in exhaust were chosen as main input variable, for they may be on-line sampled. The weights of Neural Network were optimized with simple Genetic Algorithm. The Model could be applied to predicting state variables of fungus suspended culture. Researches show that it is feasible to optimize this Neural Network system by using Genetic Algorithm.
出处
《上海生物医学工程》
1999年第4期10-12,共3页
Shanghai Journal of Biomedical Engineering
关键词
真菌深层培养
神经网络
遗传算法
Suspended Culture of Fungus Neural Network Gentic Algorithm
