Saco, Roberto (Universidad Nacional de Quilmes)

Subspace Identification of an Inverted Pendulum on a Cart Using State Variables Transformation

Scheduled for presentation during the Regular Session "System Identification" (FrA1MI), Friday, August 23, 2019, 11:20−11:40,

5th IFAC International Conference on Intelligent Control and Automation Sciences, August 21-23, 2019, Queen’s University Belfast, Northern Ireland

This information is tentative and subject to change. Compiled on April 23, 2024

Abstract

Pendulum and inverted pendulum are excellent test systems for linear and nonlinear control theory. Moreover, inverted pendulum systems are relatively easy to construct or available commercially. In this paper a linear state-space model for an inverted pendulum on a cart along with a DC motor is obtained by means of a subspace identification. The presented method require a 360-degree inverted pendulum system. We propose two white-box models of the system. One model is for the upward position and the other is for the downward position. Based on these models a parameter relationships between the two positions can be established. Then, we obtain a black-box model of the downward position using a subspace identification algorithm. The fact that the downward position of the pendulum is an open-loop stable system, allows us easily to get input-output measurements. The identified black-box model is not necessarily in the same physical state-space variables of the white-box models. Hence, a state variables transformation is needed in order to use the parameter relationships and convert the model from downward to upward position. We show experimental results by using a real pendulum system.