Output details
13 - Electrical and Electronic Engineering, Metallurgy and Materials
Queen's University Belfast
Moving window kernel PCA for adaptive monitoring of nonlinear processes
Industrial processes are complex and highly non-linear, requiring adaptive and computationally intensive solutions to render accurate measurement of product quality. The proposed kernel PCA method presents a radical solution for nonlinear industrial processes. This paper helped leverage additional funding as part of an EPSRC Science Bridge (EP/G042594/1) international technology transfer award in 2009 to develop cross-disciplinary research in related applications. Cited references and implementation of this work since publication indicates significant and broad cross-disciplinary adaptation of the proposed method, particularly in chemometrics and industrial process solutions.