Sistema de navegação inercial com controle em tempo-real para um robô autônomo trekking

Abstract

Mobile robotics is a research area that covers the control of autonomous or semi-autonomous vehicles. Among the autonomous vehicles there is a category for robotics competition called Trekking, and in this sense, the objective of this work is the development and implementation of a reliable and accurate navigation control system, based on the use of a Inertial Measurement Unit and a rotating encoder sensor for obtaining data and application in a terrestrial autonomous robot Trekking vehicle. The vehicle aims to complete a circuit, passing through predefined milestones during its journey. This route was made on a lawn with irregular terrain. For this, several sensors are used to locate the vehicle in the middle of the field, to obtain information such as direction, acceleration, speed and displacement from the inertial sensor and rotating encoder. A color sensor was used to identify the milestones. Geometric characteristics were obtained and a control system was implemented for trajectory tracking, based on the measurements of the inertial sensor and encoder. Obtaining the model for the control system was made over a robot vehicle that is the type of a conventional car, which the rear wheels are fixed and the front wheels govern the steering (Ackermann model). The proposal of this work was to use sensor signals for mapping trajectories for an autonomous vehicle controlled automatically without a human conductor, its decisions are made in a system boarded with sensors and actuators. The accuracy and reliability in obtaining data for navigation in such systems were essential. Tests and simulations that validate the operation of the control system and designed model are demonstrated. As a result, there is a standalone vehicle capable of traveling a predefined path and achieving objective milestones efficiently.