EKF-SLAM statistical bounds considering intermittent measurements

Abstract

This paper presents a theoretical study of intermittent measurement in EKF-SLAM(Simultaneous Localization and Mapping) Problem. We propose the analysis of FIM(Fisher Information Matrix) to illustrate the uncertainties statistical bounds whenever measurement data is not arrived to the system. The FIM explains the behavior of information when a measurement data is partially unavailable and therefore its existence is important to describe the system performance. The information obtained from the updated state error covariance demonstrates that the resultant state error covariance never exceeds this boundaries if a measurement data is missing during mobile robot observations. Simulation under certain conditions consistently assures the results is agreeing with the proposed analysis.