ISBN 978-3-8439-4017-7

978-3-8439-4017-7, Reihe Elektrotechnik

Sergio Alberto Flores Contreras
A Hybrid Ultrasonic System for Active-Beacon and Passive Landmark Detection Applied to Mobile Robot Localization

130 Seiten, Dissertation Universität Erlangen-Nürnberg (2018), Softcover, A5

Zusammenfassung / Abstract

This work presents the array-based ultrasonic sensors and the methods developed to create an innovative localization system for mobile robots used indoors. The system combines localization with a local positioning system (LPS), also referred in the literature as active-beacon-based localization, and landmark-based localization, to create a hybrid solution. For positioning-system-based localization, this work proposes the Range-Bearing-based Ultrasonic Positioning System (RB-UPS). The RB-UPS uses time of arrival (ToA) and angle-of-arrival (AoA) measurements to find locations with as few as one mobile station and one base station. Landmark-based localization is achieved in this work using the Hough sensor (HS). This is an echolocation sensor that applies the Hough transform to raw ultrasonic sensor data (echo profiles) for detecting geometric features, predominantly walls, in the environment of a mobile robot. The features detected by the HS are used as landmarks for localization. Hybrid mobile robot localization is achieved employing sensor data fusion methods. An extended Kalman filter (EKF) is implemented to track the position of a mobile robot using data from the RB-UPS, the HS, and the robot’s dead-reckoning sensor. The design and development, of the hardware and software of the RB-UPS and the HS, and their integration into an overall system are presented in this work. The measuring campaigns preformed for this work show that good positioning results are achieved, even inside a human environment favorable to multipath reflections that could affect sensor performance.