Multi-robot SLAM via Information Fusion Extended Kalman Filters

This paper is concerned with Simultaneous Localization and Mapping (SLAM) problem with multiple mobile robots. Each robot detects landmarks and other robots, and estimates their positions by the extended Kalman filters. To achieve good estimation accuracy, an optimal information fusion technique is adapted to the multi-robot SLAM problem. This technique involves the minimization of the estimation error covariance by weighted averaging of the state estimates from the extended Kalman filters. Simulation and experimental results are included to show the effectiveness of the present method.