PLS-Based Fault-Relevant Reconstruction
In this paper,a PLS-based fault-relevant reconstruction method is proposed for fault detection and identification.The PLS-based proposed method finds a set of latent variables to project the training data onto a process space to analyze the problem of reconstruction and is used to find the directions which can best characterize the fault effects relevant to normal status.According to the fault distribution directions,the PLS-based proposed method can eliminate the fault cause and bring faulty statistics under the control limits.The PLS-based proposed method is compared with partial least squares(PLS)to show the feasibility of the novel method,and is applied to two simulated processes.One is a simulated linear process and the other is the penicillin fermentation process.The simulated results show the improvement of the PLS-based proposed method in fault detection and identification.
Partial least squares (PLS) PLS-based fault-relevant reconstruction method Simulated linear process Penicillin fermentation process
Zhang Yingwei Tang Nan
Key Laboratory of Integrated Automation of Process Industry,Ministry of Education,Northeastern University,Shenyang,Liaoning 110004,China
国际会议
长沙
英文
4258-4263
2014-05-31（万方平台首次上网日期，不代表论文的发表时间）