Data-Driven Incipient Fault Detection via Canonical Variate Dissimilarity and Mixed Kernel Principal Component Analysis

Incipient fault detection plays a crucial role in preventing the occurrence of serious faults or failures industrial processes. In most processes, linear, and nonlinear relationships coexist. To improve performance, both linear features should be considered simultaneously. this article, novel hybrid linear-nonlinear statistical modeling approach for data-driven incipient is proposed by closely integrating recently developed canonical variate dissimilarity analysis mixed kernel principal component (MKPCA) using serial model structure. Specifically, (CVA) first applied to estimate variables (CVs) from collected process data. Linear are extracted estimated CVs. Then, (CVD) which quantifies residuals CVA state-subspace calculated explore features, components as through performing MKPCA on CVD. Fault indices formed based Hotelling's T 2 well Q statistics features. Moreover, density estimation utilized determine control limits. The effectiveness method demonstrated comparisons with other relevant methods via simulations closed-loop continuous stirred-tank reactor process.