Magnetic-map-matching Navigation For Odometry Using Low-cost Mobile Robot

Abstract

This research proposes an odometry-free mobile robot system that calculates navigation information based on dead reckoning using a control signal and a robot's dynamic model to construct the system at a minimal cost. The navigation system using control input has errors due to uncertainties such as road conditions of the operating environment, wheel slip, and inaccuracy of the control cycle. In this study, a position using magnetic map matching compensates for the cumulative error, and the yaw is measured by the gyroscope for the orientation angle error. The performance improvements are confirmed via simulations.