An approach to active simultaneous localization and mapping for mobile robot

Abstract

This paper investigates the active simultaneous localization and mapping (SLAM) for mobile robot in unknown environment. Based on the estimation by extended Kalman filter (EKF), we convert the active SLAM into a problem of multi-objective optimal control. The robot chooses the control inputs that optimize the objective function such that it can explore the environment by active, intelligent and adaptive motion behavior. Then, the above approach is extended to deal with the multi-robot active SLAM such that the map building can be finished accurately, efficiently and robustly. At last, a set of simulations is presented to show the effectiveness of our approach.