# Agile Manoeuvres using Model Predictive Control

**Kristína Fedorová, Peter Bakaráč, Michal Kvasnica ^{ *}**

*Faculty of Chemical and Food Technology, Slovak University of Technology, Radlinského 9, 812 37 Bratislava*

**E-mail:** ^{* }michal.kvasnica@stuba.sk

**Abstract:** This paper shows how model predictive control (MPC) can be used to perform agile manoeuvres in a pendulum-on-a-cart system, which is an abstraction of many mechanical systems commonly used in the industry, such as cranes. Specifically, the problem of moving a cart on which a pendulum is mounted using a free joint is rapidly moved from one position to another one while mitigating the swings of the pendulum. To achieve this goal, an optimization-based MPC strategy is employed that selects the control moves while minimizing the chosen cost function and, simultaneously, enforcing constraint satisfaction. As the controlled system is nonlinear, two options are considered. The first one solves the nonlinear MPC problem in an approximate fashion using the so-called random shooting approach. The second method is based on the first one approximating the nonlinear system by a linear one, followed by applying convex MPC techniques. The performance of both strategies was compared by means of real-time experiments.

**Keywords:** nonlinear model predictive control, convex model predictive control, random shooting, linearization

*Full paper in Portable Document Format:* acs_0336.pdf

Acta Chimica Slovaca, Vol. 12, No. 1, 2019, pp. 136—141, DOI: 10.2478/acs-2019-0019