880
Views
7
CrossRef citations to date
0
Altmetric
Articles

Nonlinear chemical processes fault detection based on adaptive kernel principal component analysis

&
Pages 350-358 | Received 06 Feb 2020, Accepted 08 May 2020, Published online: 21 May 2020

Figures & data

Figure 1. AKPCA monitoring flowchart.

Figure 1. AKPCA monitoring flowchart.

Figure 2. Monitoring results of KPCA and AKPCA for case 0 in TE process.

Figure 2. Monitoring results of KPCA and AKPCA for case 0 in TE process.

Figure 3. Monitoring results of KPCA and AKPCA for case 4.

Figure 3. Monitoring results of KPCA and AKPCA for case 4.

Figure 4. The mutation probabilities of KPCs at four sample points for fault 4.

Figure 4. The mutation probabilities of KPCs at four sample points for fault 4.

Figure 5. The similarity between DV-KPC and other KPCs at four sample points for fault 4.

Figure 5. The similarity between DV-KPC and other KPCs at four sample points for fault 4.

Figure 6. The monitoring effect diagram of DV-KPC and NDV-KPCs for fault 4.

Figure 6. The monitoring effect diagram of DV-KPC and NDV-KPCs for fault 4.

Table 1. The serial numbers of AKPCs at 150, 200, 400, and 800 sample points.

Figure 7. The monitoring results of KPCA and AKPCA for Case 19.

Figure 7. The monitoring results of KPCA and AKPCA for Case 19.

Figure 8. The monitoring results of KPCA and AKPCA for Case 21.

Figure 8. The monitoring results of KPCA and AKPCA for Case 21.

Table 2. The comparison of missed alarm rates of KPCA, SKPCA, and AKPCA in TE process.

Table 3. The comparison of false alarm rates of PCA, KPCA, SKPCA, and AKPCA in TE process.