An improved DMC algorithm used in measurement failure and it's application for paper machine control

Abstract

This paper presents an improved dynamic matrix control (DMC) algorithm that is suitable for measurement failure and its application for the baseweight control of a paper machine. The main idea of the refined DMC algorithm is to replace the output of a plant with the model output of DMC when measurement failure happens, that is, to control the model by using DMC. To achieve the aim of accurate prediction of the plant output when measurement failure happened, the standard model based prediction in DMC algorithm is improved by replacing the open loop prediction with a close loop prediction. This proposed algorithm has been applied to baseweight control of a paper machine, and the results show that it can process the measurement failure caused by paper breakage in the production process. >