Practical experiences with Python and Linux RT at the SUPSI Laboratory

Abstract

This paper presents a Computer Aided Control System Design (CACSD) environment completely based on the programming language Python. Students can perform all the control design tasks (modeling, identification, controller design, simulation) in this environment, and at the end, they can automatically generate RT code for targets like a PC or a Raspberry PI with a Linux RT OS. A Python package for control purposes have been developed at the Caltech by Richard Murray, with the aim of integrating the most used functions and methods available in other commercial and non commercial software. Using this package it is possible to design different kinds of controllers for laboratory plants (PID, state-feedback, LQR, including full and reduced order observers). An additional package developed at SUPSI allows representing the full system in a graphical block diagram, similar to xCos or Simulink, and automatically generating code for RT targets. This application offers the most used blocks required for the design of a control system. New blocks can be easily integrated into this SW. This environment has been validated at the SUPSI laboratory on different electromechanical plants: the classical Inverted Pendulum, the Disks and Spring system, the Ball on Plate and the Ball on Wheel plants.