Algorithms analysis of mobile robot SLAM based on Kalman and particle filter

Abstract

Many algorithms of mobile robot SLAM (Simultaneous Localization and Mapping) have been researched at present, however, the SLAM algorithm of mobile robot based on probability is often used in the unknown environment. In this paper, two kinds of SLAM algorithms based on probability are analyzed and compared. One kind is the SLAM algorithm based on Kalman filter: extended Kalman filter SLAM (EKF-SLAM) and unscented Kalman filter SLAM algorithm (UKF-SLAM). Another kind is the SLAM algorithm based on Particle filter: FastSLAM and unscented FastSLAM (UFastSLAM) algorithm. The difference from the four algorithms of SLAM is illustrated in terms of principle and calculation accuracy. Finally, the simulation results show that the UFastSLAM algorithm is superior to other algorithms in robot path and landmark estimation.