Study on the soft sensor and control scheme for an industrial azeotropic distillation column

In this paper, the control problem for an industry azeotropic distillation column is presented. In view of the existing problem of the control system, the author first uses soft-sensing technique to construct a soft sensor of water content of the distillation column bottom, in which the soft sensor model is built based on the field data collected by DCS. For the convenience of project application, the comparatively simple regression model is chose to form the soft sensor. After that, a scheme of inferential control is design and applied to application, and then is improved according to the actual situation after a trial period. At present, the inferential control system based on soft sensor has been running smoothly over a long period of time in worksites.