Robust State Estimation of Polymerization Processes

TUPAZ PANTOJA, JHOVANY; ASTEASUAIN, MARIANO; SÁNCHEZ, MABEL

Cancún, México

Congreso; XV Simposio Latinoamericano de Polímeros - XIII Congreso Iberoamericano de Polímeros (SLAP 2016); 2016

Sociedad Polimérica de México

The problem of state estimation in nonlinear processes has been covered extensively in the past. Typical examples of traditional chemical systems in which these problems arise are polymerization processes. A widespread state estimation technique in process control is the Extended Kalman Filter. In spite of its popularity, this strategy may present problems in the case of highly nonlinear systems. The Unscented Kalman Filter (UKF) has been developed for this type of processes. It is based on a mechanism that propagates the mean and covariance of a random variable through a nonlinear transformation.1 Because the presence of outliers distorts variable estimates, robust estimators are devised that produce reliable estimates, not only when data follow a given distribution exactly, but also when this happens only approximately due to the presence of outliers.2The aims of this paper is to compare the performance of different M-estimators in the framework of the UKF when they are applied to a copolymerization process.