基于自适应加权最小二乘支持向量机的青霉素发酵过程软测量建模

发布时间：2018-03-22 10:23

本文选题：加权最小二乘支持向量机　切入点：青霉素发酵过程　出处：《南京理工大学学报》2017年01期 　论文类型：期刊论文

【摘要】：针对生化过程软测量建模过程中样本数据可能包含的测量误差对模型性能的影响,提出一种自适应加权最小二乘支持向量机(Adaptive weighted least squares support vector machine,AWLS-SVM)回归的软测量建模方法。该方法基于最小二乘支持向量机模型,根据样本拟合误差,并结合改进的正态分布赋权规则,自适应地为每个建模样本分配不同的权值,以降低随机误差对模型性能的影响;同时采用混沌差分进化—模拟退火(Chaos differential evolution simulated annealing,CDE-SA)算法对模型参数进行优化选择,以提高模型的泛化能力。仿真实验表明,AWLS-SVM模型的预测精度及鲁棒性能优于最小二乘支持向量机(Least squares support vector machine,LS-SVM)和加权最小二乘支持向量机(Weighted least squares support vector machine,WLS-SVM)。利用Pensim仿真平台的数据,将AWLS-SVM方法用于青霉素发酵过程软测量建模,获得了较好的效果。
[Abstract]:Aiming at the influence of the measurement error that the sample data may contain on the model performance during the modeling of biochemical process soft sensor, A soft sensor modeling method for adaptive weighted least squares support vector AWLS-SVM (Adaptive weighted least squares support vector SVM) regression is proposed, which is based on the least squares support vector machine model, according to the sample fitting error, and combined with the improved normal distribution weighting rule. It adaptively assigns different weights to each modeling sample to reduce the influence of random error on the model performance. Meanwhile, chaotic differential evolution-simulated annealing algorithm (chaos differential evolution simulated annealing CDE-SA) is used to optimize the model parameters. In order to improve the generalization ability of the model, the simulation results show that the prediction accuracy and robust performance of the AWLS-SVM model are superior to those of the least squares squares support vector machine (LS-SVM) and the weighted least squares support vector machine (WLS). Using the data of Pensim simulation platform, The AWLS-SVM method was applied to the soft sensor modeling of penicillin fermentation process, and a good result was obtained.
【作者单位】：福州大学石油化工学院;
【基金】：国家自然科学基金(60804027;61374133) 福州大学科研基金(FZU-022335;600338;600567) 高校博士点专项科研基金(20133314120004)
【分类号】：TQ927;TP18

【参考文献】