Disturbance observer-based model predictive control using time-varying constraints

This paper addresses a constrained model predictive control (MPC) scheme using the disturbance observer (DOB). The purpose of the paper is to make it possible to handle constraints on the total control input, which has not been taken into account in the conventional DOB-based MPC. To this end, we bring a time-varying input constraints into design and properly adjust the MPC output at every sampling instant. The proposed DOB-based MPC scheme is applied to a position control problem of a two-link manipulator, and its effectiveness is validated through an experimental evaluation.