Figure 6.83. Phase 2 variable contributions to (a) deviations from average batch behavior at the interval of 50-60% completion of IV , (b) to linearized LV2 score at the interval of 50-60% completion of IV, (c) to linearized LV3 score at the interval of 70-80% completion of IV, (d) to linearized LV4 score at the interval of 59-65% completion of IV (dashed-dotted and dashed lines).
in time, the state is holding valuable information about earlier batches i — 1, • • • , 1 for predicting future batches. Consequently, a monitoring algorithm can be developed using state variables Xi to detect undesirable behavior from batch-to-batch.
The simplest monitoring problem would be end-of-batch monitoring based on off-line final quality measurements. In this case, contains only the end-of-batch quality data. A more comprehensive monitoring problem would include both process variable data collected during the batch and end-of-batch quality variable data. In this case, Eq. (6.134) is augmented
Was this article helpful?