SLAM with Improved Schmidt Orthogonal Unscented Kalman Filter

Simultaneous localization and mapping (SLAM) is a momentous topic for robot navigation to explore uncharted environment. To enhance the accuracy and efficiency, an improved Schmidt orthogonal unscented Kalman filter (ISOUKF) based SLAM algorithm is proposed in this paper. First, based on the Schmidt orthogonal transform (SOT) sampling, a modified unscented Kalman filter (UKF) algorithm is presented. Then, an adaptive fading factor is derived using the strong tracking algorithm, and it is introduced into the prediction covariance to improve tracking ability and accuracy. Next, the Schmidt orthogonal unscented Kalman filter is improved with square root filter to raise the efficiency of SLAM algorithm. Finally, the ISOUKF algorithm is proposed to complete the robot tracking in SLAM. The proposed algorithm provides a high precision robot tracking for SLAM and decreases the computational cost to some extent. Experiment results verify the superiority of the proposed algorithm.