Using combined IMU/Stereo Vision/cooperative GNSS System for Positioning of UxV Swarms within Catastrophic Urban Scenarios

Abstract

The usage of unmanned air (UAV) and ground (UGV) vehicles for different applications has increased in recent times. Due to time efficient operation, swarms of unmanned air and/or ground vehicles (UxV) could be used to cope with large areas or in catastrophic urban scenarios. Satisfying the demand for reliable information is one of the current challenges for unmanned vehicles. The absolute and relative position of the vehicles is indispensable for acting within a swarm. Especially in urban scenarios the reliability of position information using GNSS can be degraded by environmental conditions. To achieve this purpose this paper presents a combined system of inertial measurement unit (IMU), stereo vision and cooperative GNSS system. The fusion of the data measured by these subsystems has been realized by using an Extended Kalman Filter (EKF). Additionally an algorithm for failure detection and exclusion (FDE) has been implemented to increase the combined system's performance, concerning integrity aspects and also the position accuracy by the exclusion of degraded GNSS signals for positioning. The performance of the overall system is evaluated based on real world experiments.