Diseño e implementación de tres esquemas de control: PID, LQR y Modos Deslizantes para la estabilización del péndulo invertido sobre dos ruedas de Lego Mindstorms con aplicación de un planificador de rutas mediante el algoritmo RRT

Abstract

The aim of this work is to design and implement control algorithms to stabilize a nonlinear dynamic, multivariable and unstable system represented by two-wheeled inverted pendulum (TWIP). Furthermore, a path-planning is applying to move the TWIP from an initial point to final point. The algorithms control designed are able to keep the robot in its equilibrium point while moving. The TWIP is built with Lego Mindstorms NXT. First, three control algorithms: PID, Linear-quadratic Regulator and Sliding Mode Control are implemented in order to stabilize the robot. The algorithms are compared in terms of capacity rejection of disturbances, model uncertainties and changes reference system variables, based on results one of three controllers is selected. Then, the controller selected is using to implement a path-planning based on a Rapidly-exploring Random Tree (RRT) algorithm. The points of the path-planning are sent from a computer channel via Bluetooth.