Kalman filter-based SLAM with unknown data association using Symmetric Measurement Equations

摘要

This work investigates a novel method for dealing with unknown data associations in Kalman filter-based Simultaneous Localization and Mapping (SLAM) problems. The key idea is to employ the concept of Symmetric Measurement Equations (SMEs) in order to remove the data association uncertainty from the original measurement equation. Based on the resulting modified measurement equation, standard nonlinear Kalman filters can estimate the full joint state vector of the robot and landmarks without explicitly calculating data association hypotheses.