Range sensor aided inertial navigation using cross correlation on the evidence grid

Abstract

In current state-of-the-art navigation systems combining a global positioning system (GPS) and an inertial navigation system (INS), the two subsystems are complimentary, and the combined system only works when GPS signals are available. Much research is being undertaken on methods to replace the role of the GPS in the GPS-INS systems for use in environments where the GPS signals are not available, including areas such as urban canyons, indoors, and areas where the GPS signals have been jammed. In this paper we present a method of replacing the GPS-aiding of the INS system, by using a ranging sensor such as a lidar or radar combining with an evidence grid methodology. The evidence grid is used to create three-dimensional maps based on the measurements of the ranging sensor and the inertial system. These maps are cross-correlated in time to determine how best to correct for navigation drift, with the corrections then fed back into the extended Kalman filter to aid the inertial navigation solution. Experimental results, based on lidar measurements, show that the method is capable of remarkable accuracy, able to correct errors of less than one meter over six minutes of travel.