Canonical correlation analysis (CCA) has been used for concurrent quality and process monitoring to extract multidimensional correlation structure between process and quality variables. In this paper, a new kernel concurrent CCA (KCCCA) algorithm is proposed for quality-relevant nonlinear process monitoring, which decomposes the original space into five subspaces, including correlation subspace, quality-principal subspace, quality-residual subspace, process-principal subspace and process-residual subspace. The proposed KCCCA considers the nonlinearity in both process and quality variables, and incorporates a regularization term as well for numerical robustness. In the case studies, the Tennessee Eastman process is employed to demonstrate the effectiveness of the proposed KCCCA.
|Name||Proceedings of the American Control Conference|
|Publisher||Institute of Electrical and Electronics Engineers|
|Conference||2017 American Control Conference, ACC 2017|
|Period||24/05/17 → 26/05/17|