Cross‐directional Estimation and Predictive Control of Paper Machines Using DWT

This paper proposes a novel approach for cross‐directional (CD) estimation, modeling and control on paper machines based on discrete wavelet transforms (DWT). The CD basis weight variations are approximated at various resolutions using wavelet multiresolution analysis (WMRA). The controllable component of CD variations can be extracted from the original samples by choosing a suitable threshold resolution. An acceptable response model describing the relationship between the settings of the slice‐screws to the basis weight profile is obtained. The controller synthesis, model prediction, optimization and parameter estimation are all performed in the DWT domain. The size of optimization and control problems associated with such large dimensions can be significantly reduced.