On the Efficiency of SLAM Using Adaptive Unscented Kalman Filter

In this paper, an adaptive unscented Kalman filter (AUKF) algorithm is applied to simultaneous localization and mapping (SLAM), based on adaptation of a scaling parameter. The scaling parameter is a design parameter in the unscented Kalman filter (UKF) which can improve the quality of the approximation. An adaptive method is designed to find the suitable value for the scaling parameter to improve the accuracy of estimation. It is demonstrated that the proposed methodology significantly reduces the state estimation error and improves the navigation accuracy of an autonomous vehicle. Also, it is highlighted that the computational cost is not much affected by increasing the number of observations, especially in the SLAM application in which the number of landmarks is growing through estimation. A comparison between UKF and AUKF algorithms is also provided for the SLAM application. The efficiency and the robustness of the proposed algorithm are investigated by applying noise of different orders in simulation results. In addition, non-credibility indices are used to compare the relative performance of AUKF and UKF. The results illustrate that AUKF-SLAM is more accurate than UKF-SLAM.