Abstract
FastSLAM is a well-known solution based on particle filter to the simultaneous localization and mapping (SLAM) problem for mobile robots. The performance of the conventional FastSLAM degrades over time due to the particle depletion and it needs a large number of particles to be maintained at a high level. This paper presents an improved FastSLAM method in which niche technique and particle swarm optimization are integrated into the conventional FastSLAM. By means of the multi-modal optimization, the diversity and search ability of particles are both enhanced, and the estimation performance of particle filter is improved, so that particles would be concentrated around the true state of the mobile robot, and the precision of SLAM would be enhanced. Simulation experiment results show that the improved method is effective in SLAM, and its performance is robust even in the case of only a few particles.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have