Desenvolvimento de um sistema de localização e reconstrução de trajetórias para um veículo terrestre

Abstract

The localization task of a mobile robot is one of the most important steps towards its autonomous locomotion, since a vehicle must know its geographic position must be able to avoid known obstacles and to arrive at desired places. The fulfillment of this task requires the use of noisy information provided by many types of sensors. In this work the objective is the study of sensor fusion techniques applied to the development of a localization system for an automobile. This automobile will be used later as a mobile robotic platform in UFMG. The developed system not only provides information concerning the geographic position of the vehicle, but also estimates speed and orientation. This is accomplished by the use of position, speed and acceleration sensors, which are adequately installed and calibrated. The sensors measurements are processed using estimation algorithms based on Kalman Filter. As example of application of the localization system proposed in this work, an onboard electronic guide was developed, which, based on maps stored previously on a computer, can display geographical information to the vehicle driver.