The Analysis of Covariance Matrix for Kalman Filter based SLAM with Intermittent Measurement

Abstract

This paper presents an analysis of the impact of intermittent measurement to the Simultaneous Localization and Mapping (SLAM) of a mobile robot. Intermittent measurement is a condition when the mobile robot lost its measurement data during observation due to sensor failure or imperfection of the system. This is crucial, since SLAM requires measurement data recursively for data update in estimating its current states. In this study, the analysis focused on the effect of intermittent measurement on the state error covariance matrix for two basic conditions; mobile robot is stationary and moving. The impact on the determinant of covariance matrix is observed. From the analysis, it can be concluded that intermittent measurement may lead to incorrect estimation of robot position and increment of state error covariance matrix.