Robust Least Square Filter for Simultaneous Localization and Mapping

Abstract

In this work, a robust least square filter (RLSF) based simultaneous localization and mapping (SLAM) has been presented. First a RLSF for nonlinear systems has been derived using the concept of extended Kalman filter design. For designing the filter, system uncertainties have been considered in the system model. This kind of filter may be useful in the field of robotics and fault diagnosis. Then the derived filter has been applied to SLAM for evaluating its performance and applicability. Finally the proposed method is simulated in MATLAB with numerical data and compared with EKF based SLAM. It is found that the present method is performing better than EKF based SLAM.