KAGAKU KOGAKU RONBUNSHU, Vol.36, No.4, 394-404, 2010
Model Predictive Control Based on ARXModels
This paper presentsanovel method of linear model predictivecontrol (MPC) based on ARXmodels in incremental form. Unmeasured disturbances are popular in industrial control systems and must be treated appropriately. Manystudies have been performed on nonlinear MPC and estimation of unmeasured disturbances. The proposed method provides a robust control algorithm for unmeasured disturbances in a linear MPC framework without using estimation of unmeasured disturbances. The feature of the MPC is an algorithm that can use the AR term of ARX models for unmeasured disturbance as a tuning parameter. The application and performance of the control method are illustrated by simulation of several examples.