Ant system based LMB filter for SLAM implementation in ROS platform

With a combination of ant system and labeled multi-Bernoulli(LMB)filter, a novel framework is proposed for the problem of simultaneous localization and mapping (SLAM). Within the proposed framework, the locations and the number of features can be jointly estimated and managed by random finite set. Besides, a real-time moving ant estimator (RMAE) is employed to estimate moving vehicle trajectory. Compared to the recently developed methods, the proposed algorithm uses the artificial ants to take place of particles to gather around their areas of interest through ants' positive feedback search function, and thus ensure the perfect implementation of the algorithm in robot operating system(ROS). Experimental results provide a better map as well as an improved estimate accuracy of the vehicle's trajectory for the proposed approach, and the performance is better than both the PHD-SLAM and the LMB-SLAM.