The paper describes the analysis of paper machine process datausing discrete wavelet transforms. The techniques have been adapted froma general signal analysis theory. The authors previously showed (1996)that wavelets are an effective representation for the detection of basisweight and moisture process variations in noisy data and lead toimproved estimation and visualization of the machine direction and crossmachine variations. This paper discusses data storage using the waveletrepresentation, and shows that the method also allows significantcompression of the scanned data without diminishing the accuracy withwhich profiles can be reconstructed. It is shown that the compressionmethod can be embedded into the estimation algorithm, producingexcellent results without major expense in computation time. The abilityto reduce data storage requirements is of increasing importance inmill-wide process monitoring systems and quality assurance
展开▼
