Biomass Estimation in Batch Process through Particle Filters

QUINTERO OLGA LUCÍA; A. N. AMICARELLI; F. A. DI SCIASCIO

Lima

Congreso; XV Congreso Latinoamericano de Control Automático CLCA 2012; 2012

This work proposes a biomass estimator based on nonlinear filtering techniques. The performance of the proposed estimator is illustrated in a batch process, namely the δ-endotoxin production of Bacillus thuringiensis (Bt). Biomass estimation is performed indirectly using the information of measurement concentrations of the process, namely, dissolved oxygen and the substrate. The dynamics of bioprocesses and particularly of this batch biotechnological process are generally strongly nonlinear and the nature of the uncertainties and perturbations cannot be supposed to be Gaussians distributed. For these reasons, nonlinear filtering techniques constitute an adequate option as estimation tool. The simulations? results and its experimental verifications demonstrate the acceptable performance of the proposed biomass estimator.