C. Jaeckle et J. Macgregor, PRODUCT DESIGN THROUGH MULTIVARIATE STATISTICAL-ANALYSIS OF PROCESS DATA, Computers & chemical engineering, 20, 1996, pp. 1047-1052
A methodology is developed for finding a window of operating condition
s within which one should be able to produce a product having a specif
ied set of quality characteristics. The only information that is assum
ed to be available is that contained within historical data on the pro
cess obtained during the production of a range of existing product gra
des. Multivariate statistical methods are used to build and to invert
an empirical model of the existing plant operations to obtain a window
of operating conditions that are capable of yielding the desired prod
uct, and that are still consistent with past operating procedures and
constraints. The methods and concepts are illustrated using a simulate
d high pressure tubular reactor process for producing low density poly
ethylene.