Steady-state sequential distributed parameter filtering to estimate temperature and composition profiles of a fixed-bed tubular catalytic reactor

Abstract

An optimal steady-state sequential distributed filter has been derived and tested. A limited-memory type modification is presented and shown to have superior convergence properties compared to the optimal filter when a number of replicate data sets are available for processing. Experimental data from a styrene-monomer tubular-reactor pilot plant were processed using the modified filter since replicate sets of data were available. Results show steady-state sequential filtering to be effective and robust.