首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于IRLS-ELM生物发酵在线软测量建模方法
引用本文:刘国海,张东娟,梅从立.基于IRLS-ELM生物发酵在线软测量建模方法[J].东南大学学报(自然科学版),2011,41(Z1):10-13.
作者姓名:刘国海  张东娟  梅从立
作者单位:江苏大学电气信息工程学院,镇江,212013
基金项目:国家高技术研究发展计划(863计划)资助项目(2007AA04Z179)
摘    要:为解决生物发酵过程中生物量浓度难以在线测量的问题,提出一种基于改进的最小二乘正则化极限学习机(IRLS-ELM)软测量建模方法并将其应用于红霉素发酵过程生物量浓度的在线预测中.根据误差反馈原理,将训练误差作为输入建立带反馈的神经网络,以提高模型预测精度.并将加权最小二乘法引入到ELM中改进其数学模型,削弱离群点或者不稳...

关 键 词:极限学习机  软测量  反馈输入  发酵过程

Soft sensor modeling based on improved regularized least-squares extreme learning machine method
Liu Guohai,Zhang Dongjuan,Mei Congli.Soft sensor modeling based on improved regularized least-squares extreme learning machine method[J].Journal of Southeast University(Natural Science Edition),2011,41(Z1):10-13.
Authors:Liu Guohai  Zhang Dongjuan  Mei Congli
Institution:(School of Electrical and Information Engineering,Jiangsu University,Zhenjiang 212013,China)
Abstract:In order to solve the problem that biomass concentration is difficult to directly measure in the fermentation process,a soft sensor modeling method based on improved regularized least-squares extreme learning machine(IRLS-ELM) is proposed and it is used to predict the biomass concentration in the erythromycin fermentation process.According to the principle of error feedback,the training error can be used as the input of the ELM to improve the prediction accuracy of the model.In order to reduce the effects of outliers and uncertainty,the weighted least squares method is incorporated in the model.Finally,a soft sensor model based on the IRLS-ELM is constructed for predicting in erythromycin fermentation process.The results show that the mean square error(MSE) of the IRLS-ELM soft sensor model is less than that of the ELM and the RLS-ELM with the same number of hidden nodes.In addition,the IRLS-ELM has compact structure and high stability and its error does not change significantly when the number of hidden nodes becomes fewer.Compared with the ELM and RLS-ELM model,the IRLS-ELM model has higher prediction accuracy and stronger generalization capability.
Keywords:extreme learning machine  soft sensing  feedback input  fermentation process
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号