Indoor Flight Navigation System of Micro Air Vehicle with Inertial Sensor Drift Correction

This paper describes an indoor flight control system capable of autonomously exploring indoor areas with microelectromechanical systems (MEMS). Whereas the significant drift of MEMS is generally mitigated using occasional GPS location estimates, GPS is unavailable indoors. In order to solve this problem, a Rao-Blackwellized Particle Filter SLAM algorithm was used to make MAVs localize themselves and the drift error was examined afterward. This paper develops the way to correct the sensor drift by the pose estimated by localization.