Nonlinear Dynamic Process Monitoring Using Canonical Variate Kernel Analysis

Most industrial systems today are nonlinear and dynamic. Traditional fault detection techniques show their limits because they can hardly extract both dynamic features simultaneously. Canonical variate analysis (CVA) shows its excellent monitoring performance in for processes but is not applicable to processes. Inspired by the CVA method, a novel process namely, “canonical kernel analysis” (CVKA), proposed this work. The way different from traditional canonical (KCVA). In sequential structure, new approach firstly extracts linear data through followed principal component residual space. CVKA method then applied TE case study, proving of compared other common approaches TE-like