摘要
针对生化反应过程中软测量模型存在的模型失效问题,提出了一种基于模糊C均值聚类(FCM)和动态LS-SVM的混合建模方法。首先,采用FCM算法将训练集分成具有不同聚类中心的子集,然后对每一类分别采用LS-SVM进行训练并建立子模型。对于带有新信息的样本数据首先计算其对每一类的模糊隶属度函数,然后用隶属度最大的一类所对应的子模型进行动态学习,并更新子模型。将所提出的软测量建模方法用于对L-赖氨酸发酵过程关键生物量参数的预测,实验结果表明所提出的建模方法可以有效地增强软测量模型适应工况变化的能力,提高其预测精度。
Aiming at the problem that soft-sensing model cannot be updated with the bioprocess changes,a soft-sensing modeling method based on hybrid fuzzy C-means clustering algorithm and dynamic LS-SVM is proposed.FCM is used to separate a whole training data set into several clusters with different centers,each subset is trained by LS-SVM and sub-models are developed to fit different hierarchical properties of the process.New sample data that represent new operation information are introduced in the model,and the fuzzy membership function of the sample to each clustering is first computed by FCM algorithm.Then,a corresponding LS-SVM sub-model of the clustering with largest fuzzy membership function is used to perform dynamic learning,so the model can be updated on-line.The proposed method is applied to predict the key biological parameters in L-lysine fermentation process.Simulation results indicate that the proposed method actually increases the model adaptive abilities under various operation conditions and improves its generalization capability.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2010年第2期404-409,共6页
Chinese Journal of Scientific Instrument
基金
国家863计划(2007AA04Z179)
高等学校博士学科点专项基金(20070299010)资助项目