Desarrollo de un controlador en espacio de estados para un sistema Ball and Plate realimentado por visión artificial.

Abstract

In the present project, the controller is detailed in the state space for control of a ball and plate system fed by artificial vision. To initialize the controller, we proceeded to find the mathematical model of the non-linear plant by analyzing different equations that governs the dynamics of the system, to obtain its transfer function it is necessary to linearize the equation found and represent it in variable states in order to to design a controller in state space. It is analyzed if the controller is optimal and does not present errors through tests carried out in the Matlab software, observing the trajectory results. In order for the controller to become more stable, an observer was implemented to eliminate disturbances or noise in the system. The objective of this project is to control the position of the sphere on a plane that changes its angles of inclination to position the ball at a specific point on the platform, this can be controlled by a set point by means of artificial vision that indicates the coordinates "X" and "y" in which the sphere moves. A visual interface was developed in Labview to obtain the controller's gains and to be able to choose the desired position of the ball. The calculation of error was obtained by means of the integral equation of the absolute error (IEA), implementing a performance comparison between the conventional PID controller and the controller in Statespace.