Soft sensor for compression and burst strength: identification and uncertainty analysis
Research output: Contribution to journal › Article › Scientific › peer-review
Details
| Original language | English |
|---|---|
| Pages (from-to) | 38-47 |
| Number of pages | 10 |
| Journal | Journal of Science and Technology for Forest Products and Processes |
| Volume | 7 |
| Issue number | 1 |
| Publication status | Published - 2018 |
| Publication type | A1 Journal article-refereed |
Abstract
We present a work process for developing soft sensors, and apply it for board compression and burst strength. Variables are prescreened and data preprocessed based on paper physics and process knowledge. The measurement uncertainty of the soft sensor is analyzed, including effects of the noise in laboratory strength measurements that constitute the reference in the identification of the soft sensor model, and the effects of parameter uncertainties. The uncertainty analysis provides the end users a means to assess the usefulness of the information at different operating situations, and to recognize when soft sensor predictions are not to be trusted due to, e.g. extrapolation.