Soft sensor design for continuous product quality prediction in debutanizer column

Data-driven soft sensors have been introduced as an effective solution for industrial process control corresponding to online monitoring of product quality, which always face with technical or economic limitation to measure key variables. The objective of this paper is to present a new data-driven soft sensor based on state dependent parameter models for online monitoring of butane content in the bottom product of debutanizer column, on the basis of a set of available measurements. The model parameters are assumed to be function of the system states, which are estimated by data-based modeling philosophy and state dependent parameter method. Considering statistical and sensitivity analysis, the best input variables are obtained. Two soft sensing models are provided with different number of parameters. The prediction results match the best with the actual outputs, which demonstrate the effectiveness of proposed method for quality prediction as an alternative for laboratory testing.

Keywords:

Data-driven soft sensor , state dependent parameter , quality prediction , debutanizer column , butane content