Comprehensive monitoring of industrial processes using multivariable characteristics evaluation and subspace decomposition,The Canadian Journal of Chemical Engineering

当前位置： X-MOL 学术 › Can. J. Chem. Eng. › 论文详情

Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)

Comprehensive monitoring of industrial processes using multivariable characteristics evaluation and subspace decomposition
The Canadian Journal of Chemical Engineering ( IF 2.1 ) Pub Date : 2021-07-18 , DOI: 10.1002/cjce.24265
Shuai Li _{1,

2,

3,

4} , Xiaofeng Zhou _{1,

2,

3} , Haibo Shi _{1,

2,

3} , Fucheng Pan _{1,

2,

3} , Xin Li _{1,

2,

3} , Yichi Zhang _{1,

2,

3}

Affiliation

Gaussianity, non-Gaussianity, linearity, and nonlinearity generally coexist within industrial process variables, and should be taken into account simultaneously for process modelling with monitoring. This paper presents a comprehensive monitoring method of industrial processes using multivariable characteristics evaluation and subspace decomposition. First, a multivariable characteristics evaluation method is presented to divide the process variables into the Gaussian linear, Gaussian nonlinear, non-Gaussian linear, and non-Gaussian nonlinear subspaces. Second, the PCA-ICA-KPCA-KICA-based multivariable subspace decomposition is proposed for process modelling. Furthermore, comprehensive monitoring is developed and final results are combined using comprehensive statistics. By multivariable characteristics evaluation and subspace decomposition, the proposed method could evaluate and seek the multivariable characteristics and enhance the performance of process monitoring. The effectiveness and feasibility of the proposed comprehensive monitoring method are demonstrated by a numerical system and the benchmark Tennessee Eastman (TE) process.

更新日期：2021-07-18

点击分享查看原文

点击收藏

阅读更多本刊最新论文

全部期刊列表>>