A new estimation method by utilizing on-line tracking simulator

In the chemical process, it is hard to describe the polymer property by a rigorous model based on the physical and chemical theory. The soft sensor with the statistical data processing is often used for modeling to monitor and to control the polymer quality. We confirmed regarding the estimation accuracy that using data from the on-line tracking simulator is better than using data from the conventional use of the soft sensor. The on-line tracking simulator perfectly simulates the target and provides the virtual input data which can not be measured by the soft sensor on the real time. We propose the application of the rigorous and the statistical model to the plant operation.