Near time-optimal model predictive control using an L1-norm based cost functional

A comparison between standard model predictive control (MPC), an L1-norm based MPC formulation and time-optimal control is presented in this work. It is shown that MPC with an L1-norm based cost functional can successfully approximate time-optimal control with the advantage of a computationally less complex problem formulation. Furthermore, compared to standard quadratic form based MPC, the L1-norm based method shows less overshoot in the step response. The compared approaches are validated by measurements for a highly dynamical permanent magnet DC motor which forms a two-mass system with a load.