Model Predictive Control of DC Servomotor using Active Set Method

This paper presents the design and real-time hardware implementation of Active Set Method (ASM) based Linear Model Predictive Controller (MPC) for position control of DC Servomotor. A key component of MPC is to solve an online convex Quadratic Programming (QP) optimization problem at each sampling instance within specified sample period. In view of ASM, Cholesky factorization based linear solver is used to solve the KKT system, which renders faster execution of MPC. The designed approach is demonstrated for real-time MPC implementation on Quanser QET DC Servomotor setup.