Calibration and measurement processing for ultrasonic indoor mobile robot localization systems

This work introduces an indoor localization system for mobile robots using Time-of-Flight measurements between an ultrasound signal emitter mounted on the robot and ultrasonic sensors (anchors) mounted over the workspace of the robot. In the first part of the study a calibration method was proposed for Time-of-Flight measurement based localization systems. The influence of the anchor's position on localization performance was analyzed to show the necessity of precise calibration. The calibration procedure was formulated as an optimization problem and its efficiency was showed through real measurements. In the second part of the study an Extended Kalman Filter based predict - update type estimator was proposed to increase the measurement rate of the localization system. For the state estimator an omnidirectional robot model was developed which inputs are provided by an Inertial Measurement Unit. Experimental results using a KUKA youBot mobile robotic platform are also presented to show the applicability of the proposed indoor localization method.