Fast Large-scale Slam With Improved Accuracy In Mobile Robot

Simultaneous localization and mapping (SLAM) is a hot topic of intelligent robotics, because it is essential for autonomous navigaition of mobile robots. Extended Kalman Filter (EKF) is one of the most popular algorithms for sloving SLAM problem according to the relative researches. The filter provides nonlinear motion model and measurement model for perception step and correction step of estimating the state vector and its covariance. The process of estimating is fast and the results are precise for small-scale invironment applications. However, the applications of SLAM algorithms are recently extented to outdoors where large-scale is one of main characteristics. The standard EKF SLAM has two important limitations. High computaional cost O ( n 3) and error accumulation aroused by linearization will limit the use of the standard EKF SLAM in large-scale environments.In order to solve these problems, however, at the same time reserve the advantages of EKF in the applications of small-scale environments. A Fast Map Joining (FMJ) SLAM algorithm is proposed in this paper to achieve the global localization and mapping of mobile robot in the large-scale environments simultaneously. The FMJ SLAM algorithm divides the global map into a sequence of local sub-maps whose sizes are determined according to the density of the landmarks in the environment. The final localization and mapping is achieved once the sub-maps are jointed accordingly. The proposed algorithm can efficiently improve the accuracy of the SLAM and reduce the computational load compared with the standard EKF SLAM. Simulations are conducted to validate the proposed technique.