Consideration of Soft Sensor Methods Based on Time Difference and Discussion on Intervals of Time Difference

Abstract

In chemical plants, soft sensors are widely used to estimate process variables that are difficult to measure online. The predictive accuracy of soft sensors decreases over time because of changes in the state of chemical plants, and soft sensor models based on time difference (TD) have been constructed to reduce the effects of deterioration with time, such as drift. However, many details of models based on TD remain to be clarified. In this study, TD models are discussed in terms of noise in data, auto-correlation in process variables, and degree of model accuracy, among others. We theoretically clarify and formulate the differences of predictive accuracy between normal models and TD models. The relationships and the formulas of TD were verified by analyzing simulation data. Furthermore, we analyzed data obtained by dynamic simulation of an existing full-scale depropanizer distillation column and examined various TD intervals and the predictive ability of TD models. It was confirmed that the predictive accuracy of TD models increased when TD intervals were optimized.