LAUSR

The FastSLAM algorithm has become an effective way to solve the simultaneous localization and mapping (SLAM) problem. However, measured in terms of the number of particles required to build an accurate map, currently, its accuracy cannot be easily enhanced because of particle degeneracy. In view of these problems, in this paper, we present a fast algorithm of SLAM based on the ball particle filter (Ball-PF), which originates from the modification of the box particle filter (Box-PF). First, the transform relationship between Box-PF and Ball-PF are studied in depth so as to show the advantages of Ball-PF with respect to solving the interval constraints satisfaction problem and prevent from breaking down effectively. Then, a new fast algorithm of SLAM is designed with Ball-PF, in which the firefly algorithm is used to maintain the diversity of the ball particles to increase the consistency of the pose estimation effectually. Furthermore, the map matching technique is used to compute the weight of the ball particles and learn the grid maps incrementally. The simulation and experimental results demonstrate the performance superiority of the proposed algorithm.