带时延估计的加权高斯模型软测量建模  

Weighted Gaussian model soft sensor modeling and its application with time delay estimation

作　　者：车笑卿 熊伟丽[1,2] Che Xiaoqing;Xiong Weili(School of Internet of Things Engineering,Jiangnan University,Wuxi 214122,China;Key Laboratory of Advanced Process Control for Light Industry（Ministry Of Education）,Jiangnan University,Wuxi 214122,China)

机构地区：[1]江南大学物联网工程学院,无锡214122 [2]江南大学轻工过程先进控制教育部重点实验室,无锡214122

出　　处：《仪器仪表学报》2018年第9期195-202,共8页Chinese Journal of Scientific Instrument

基　　金：国家自然科学基金(61773182)项目资助

摘　　要：针对实际化工过程存在的时延及随机噪声的干扰问题,提出一种改进的在线软测量方法。该方法首先采用滑动灰关联度分析法估计过程时延参数,提取过程时延信息;当新样本到来时,基于离线阶段估计的时延参数对建模样本进行重构,并通过新样本相对于训练样本的权重建立加权高斯模型,构建输入输出变量的联合概率密度函数;最后,通过条件分布函数实时估计输出变量值。通过数值仿真和工业脱丁烷塔过程丁烷浓度的在线估计,验证了所提方法的有效性。Aiming at the delay and random noise interference problems existing in actual chemical processes, an improved online soft-sensor method is proposed. The method firstly uses the moving gray relational analysis method to estimate the process delay parameter and extract the process delay information. When the new sample arrives, the modeling sample is reconstructed based on the time delay parameter estimated in off line phase. According to the weight of the new sample relative to the training sample, a weighted Gaussian model is established, and a joint probability density function of input and output variables is constructed. Finally, the output variables are estimated in real time based on the conditional distribution function. The effectiveness of the proposed method is verified through a numerical simulation example and the online estimation of the butane concentration in industrial debutanizer column process.

关 键 词：变量时延 滑动灰关联度分析 加权高斯模型 在线建模 

分 类 号：TH7[机械工程—仪器科学与技术]