摘要
提出一种改进的BP神经网络(IBPNN)用以建立发酵过程中菌体浓度软测量模型.结合菌体浓度变化范围大这一特点,将传统BP网络的误差函数进行了改进,并利用最优停止法对网络进行训练,避免了过拟合现象.最后针对诺西肽发酵过程中菌体浓度的估计问题,根据隐函数定理选取辅助变量,应用IBPNN建立菌体浓度软测量模型,实验结果验证了所提方法的有效性.
An improved BP neural network （IBPNN） is presented to develop a soft sensor model of biomass in fermentation processes. Combined with the characteristic that biomass can vary in a wide range, the error function of the traditional BP network is improved. Meanwhile, optimal stopping rule is used to avoid over-fitting. According to the estimation of biomass in Nosiheptide fermentation process, the secondary variables are selected according to the implicit function existence theorem, and then a soft sensor model of biomass is developed by using the IBPNN, The testing result shows the effectiveness of the presented approach.
出处
《控制与决策》
EI
CSCD
北大核心
2008年第8期869-873,878,共6页
Control and Decision
基金
国家973计划项目(2002CB312201)
关键词
软测量
神经网络
算法改进
辅助变量选取
发酵
Soft sensor
Neural network
Algorithm improvement
Selection of secondary variables
Fermentation
