Robust trilateration based indoor localization method for omnidirectional mobile robots

摘要

This paper proposes an enhanced trilateration based indoor localization scheme for omnidirectional mobile robots. The method assumes that the motion of the robot is followed by a sensor network the elements of which can determine the robot-sensor distance based on Time of Flight measurements. Based on multiple distance measurements the sensor network can compute the robot's position using trilateration. This study proposes a robot position computation method for the case when only one sensor is active for a limited period of time by applying augmented polynomial regression for robot path estimation. For fast localization a sensor fusion method is also proposed which combines the sensor network's position measurements with the measurements of an Inertial Measurement Unit placed on the robot. The sensor fusion takes into consideration the characteristics of the omnidirectional robot models. Simulation measurements and experimental results using a KUKA youBot mobile robotic platform are provided to show the applicability of the proposed indoor localization methods.