A Consistent Union-for-Fusion Approach to Multi-Robot Simultaneous Localization and Target Tracking

This paper provides a fully decentralized approach for multi-robot simultaneous localization and target tracking based on extended Kalman filter and covariance union (CU), referred to as (EDMR-SLTT). In the proposed approach, each robot maintains the latest estimate of itself and targets, and information exchange only takes place between two robots when they obtain relative measurements of each other. Moreover, we have proved that when CU is used to fuse the target state estimate with the target state estimated by teammate robots, the positive definiteness of the robot and target joint covariance matrix is guaranteed without any calculation of the robot-to-target correlation terms. Finally, simulation and experimental results have shown that the EDMR-SLTT approach is superior to alternative state-of-the-art approaches with comparable processing and communication costs.