Biomedical Engineering Reference
In-Depth Information
Figure
12.10. Contribution plot (top) and actual variable time series charts (bottom) for
variables 1 and 2 against their mean trajectories and
3 SD variability ranges.
“relevant time,” should be smooth, continuous, monotonically increasing (or decreas-
ing), nonnoisy, and spanning the range of all other process variables measured within the
same unit procedure of interest [65]. A practical application of this approach in
biopharmaceutical purification processes is proposed by Larson et al. [37] for monitoring
