Multi-robot map building in unknown environment using cooperative correction

This paper focuses on the multi-robot simultaneous localization and map building in unknown environment (SLAM) problem. Cooperative correction approach is proposed to take the advantage of multi-robot cooperative exploration and improve the map building quality. The cooperative correction contains weak cooperative correction and strong cooperative correction. Both of them are able to decrease the uncertainty using Extended Kalman Filter (EKF). This paper also introduced three goals related to different cooperative correction fashion. The robot moves with different goals in different situations. The transition between different goals depends on the current map building quality. Simulation shows the effectiveness of the proposed approach.