Underground mine navigation using an integrated IMU/TOF system with unscented Kalman filter

Abstract

Platform pose (localization and orientation) information is a key requirement for autonomous mobile systems. The severe natural conditions and complex terrain of underground mines diminish the capability of most pose estimation systems, especially GPS. Our research interest is focused on using low-cost off-the-shelf IMU to improve the Active Beacon Positioning System (ABPS) developed here at the CSIR. This paper proposes a novel pose estimator, for underground mines, that fuses together data from the ABPS and low-cost MEMS based IMU. This pose estimator uses a square-root unscented Kalman filter (SR-UKF) to fuse the data together. The method is evaluated by building a complete system in a lab.