The application of GPC and its parallel implementation for cross-directional basis weight profile control in paper machines

This paper applies the multivariate generalized predictive control (GPC) strategy and its parallel algorithm for the cross-directional (CD) basis weight profile in paper making processes. The coordinative control of multiple actuators is realized. The CARIMAX model, which represents the CD weight profile response to the slice lip position settings, has been established. According to the model structure, the system is decomposed into several sub-systems. The GPC algorithm is distributed in every sub-system and the parallel control algorithm is obtained. It is proposed that this parallel control algorithm is to be implemented in a multiprocessor system. Computer simulation results indicate that the control strategy is effective and practical.