CIESC Journal ›› 2008, Vol. 59 ›› Issue (10): 2553-2560.
Previous Articles Next Articles
soft sensor; selection of secondary variables; weighted; RBF neural network; biomass; fermentation
Online:
Published:
杨强大;王福利;常玉清
Abstract:
Combined with the practical situation of Nosiheptide fermentation process, a weighted radial basis function neural network (WRBFNN) based biomass soft sensor modeling method is presentedBased on the unstructured model of Nosiheptide fermentation process, the secondary variables were selected according to the implicit function existence theorem, which made the selection theoretically strictAccording to the characteristics that biomass could vary in a wide range, the error function of the traditional RBFNN was improvedEach batch sample was selfadaptively weighted according to their predicting ability to the predicted object, and then WRBFNN was used to develop the biomass soft sensor modelThe testing results showed the effectiveness of the presented approach.
Key words: 软测量, 辅助变量选择, 加权, RBF神经网络, 菌体浓度, 发酵
软测量,
摘要:
结合诺西肽发酵过程的实际情况,提出了基于加权RBF神经网络(weighted RBF neural network, WRBFNN)的菌体浓度软测量建模方法。在诺西肽发酵过程非结构模型的基础上,根据隐函数存在定理确定出辅助变量,从而使其选择有严格的理论依据。针对菌体浓度变化范围大这一特点,将传统RBF神经网络(RBF neural network, RBFNN)的误差函数进行了改进;然后根据每批训练样本对被预测对象的预估能力,自适应地为各个批次的训练样本分配权重,进而实施WRBFNN建模。实验结果验证了所提方法的有效性。
关键词: 软测量, 辅助变量选择, 加权, RBF神经网络, 菌体浓度, 发酵
soft sensor, selection of secondary variables, weighted, RBF neural network, biomass, fermentation. Weighted RBF neural network based soft sensor of biomass in Nosiheptide fermentation process[J]. CIESC Journal, 2008, 59(10): 2553-2560.
杨强大, 王福利, 常玉清. 基于加权RBF神经网络的诺西肽发酵过程菌体浓度软测量 [J]. 化工学报, 2008, 59(10): 2553-2560.
0 / / Recommend
Add to citation manager EndNote|Ris|BibTeX
URL: https://hgxb.cip.com.cn/EN/
https://hgxb.cip.com.cn/EN/Y2008/V59/I10/2553
LI Jiubao;LIU Xinggao
Melt index prediction based on PSO_SA algorithm
GENG Zhiqiang;ZHU Qunxiong;GU Xiangbai;LIN Xiaoyong
Optimal control of cracking depth based on multi-swarm competitive PSO-RBFNN for ethylene cracking furnace
Soft sensor modeling based on DE-LSSVM
LI Zhe;TIAN Xuemin
Soft sensor modeling method based on secondary variables KNN analysis
Modeling for size distributions of flocculating yeast cell flocs during continuous ethanol fermentation
CHEN Guo; ZHAO Ya’nan;HUANG He;YAO Shanjing
1,3-Propanediol production from glycerol by Klebsiella pneumoniae encapsulated in NaCS/PDMDAAC capsules
YAN Xuefeng;YU Juan;QIAN Feng
Development of naphtha dry point soft sensor by adaptive partial least square regression
CHENG Zhong;CHEN Dezhao
WBRPLSR method and its application to dynamic chemical process modeling