摘要：AbstractThis paper demonstrates the application of the linear parameter-varying (LPV) framework to control a copolymerization reactor. An LPV model representation is first developed for a nonlinear model of the process. The LPV model complexity in terms of the model order and the number of scheduling variables is then reduced by truncating those system states that have insignificant direct influence on the input-output behavior of the system and do not directly appear in the output equations of the model. It is important to note that these truncated states are still preserved in the reduced model by affecting the scheduling parameters and hence enabling the representation of the input-output map. Using the derived model, a linear fractional transformation (LFT) based LPV controller synthesis approach is used to synthesize a controller for the process. Simulation based studies of the closed-loop behavior of the system regulated by the designed LPV controller demonstrate that the LPV controller solution outperforms a model predicative control designed previously for this system in terms of the achieved control performance and the online computational effort.
