Adrian Martin and M. Reza Emami
Distributed control, multi-robot system, particle filter, simultaneous localization and mapping
A new approach to simultaneous localization and mapping (SLAM) using particle filters has been developed to address the issue of limited and changing processing resources in autonomous exploration tasks. This algorithm is able to store and integrate historical data after the fact to improve current estimates and to delay weight processing until it is required. The foundation of the approach is developed here and is proposed as a new strategy that can complement many existing particle filter SLAM algorithms. The algorithm is demonstrated in a Cooperative SLAM implementation with an occupancy grid map. Simulation results show a 17% reduction in exploration time compared to two traditional SLAM approaches.
Important Links:
Go Back
