Fault Detection Method for Non-linear Systems Based on Kernel Independent Component Analysis and Support Vector Data Description

Complex industrial process data have the features of non-Gaussian and strong non-linear, so we propose a fault detection method for non-Gaussian and non-linear systems based on the kernel independent component analysis and the support vector data description (KICA-SVDD). Firstly, we apply the KICA method to extract the features of the data. Then we use the SVDD to model the extracted leading independent component and to calculate the statistics and the control limits. So that the faults on the non-Gaussian and non-linear system can be detected. Finally, the experimental results on the Tennessee-Eastman (TE) process's simulation study show that the proposed method reduces the fault misclassification ratio and the miss detection ratio, which verifies the proposed method's feasibility and validity.