Thermodynamic diagram based estimation structure design for ternary distillation column

Adjustable-structure design for ternary distillation columns

“…For instance, the EKF has a data assimilation mechanism with innovation injection over the entire set of model states, and the questions are whether such complete innovation mechanism is actually at play, how many temperature sensors should be used and where they should be located, and if the measurement structure should be adaptive and how this should be done. In our previous studies, the innovated/non-innovated state partition concept (Alvarez et al, 2000;Lopez et al, 2004) has been applied to ternary distillation columns (Pulis et al, 2006a), and thermodynamically interpreted (Pulis et al, 2006b), establishing the feasibility of drawing simplified estimation algorithms. Here, a further step is taken along this structure-oriented geometric framework, with emphasis on: (i) regarding the estimation structure (i.e., the innovated/noninnovated partition, number of sensors and their location) as a design degree of freedom, (ii) finding out to which extent the estimator functioning depends or not on the choice of (geometric or EKF) estimation algorithm, and (iii) exploring the feasibility of performing the data assimilation task in the light of an adaptive sensor location scheme.…”