An Enhanced CSLAM for Multi-robot based on Unscented Kalman Filter

An Enhanced CSLAM for Multi-robot based on Unscented Kalman Filter

This paper proposes an unscented Kalman filter (UKF) based coordinative, simultaneous localization and mapping (CSLAM) system, in which robots share common mapping information. The SLAM information obtained by a master robot is shared with slave robots, which estimate only their own localizations using comparatively simple sensors. The behavior of the slave robots depends on the reconstructed CSLAM using information transmitted by the master robot. The proposed process re-duces the processing burden of the slave robots, which results in a reduction of the calculation time and the complexity of their hardware system. By comparing the proposed algorithm with some conventional methods in terms of system stability, the efficiency of the proposed method is verified.

This paper proposes an unscented Kalman filter (UKF) based coordinative, simultaneous localization and mapping (CSLAM) system, in which robots share common mapping information. The SLAM information obtained by a master robot is shared with slave robots, which estimate only their own localizations using comparatively simple sensors. The behavior of the slave robots depends on the reconstructed CSLAM using information transmitted by the master robot. The proposed process re-duces the processing burden of the slave robots, which results in a reduction of the calculation time and the complexity of their hardware system. By comparing the proposed algorithm with some conventional methods in terms of system stability, the efficiency of the proposed method is verified.