Cooperative Visual-Range-Inertial Navigation for Multiple Unmanned Aerial Vehicles

In this article, the cooperative navigation issue is investigated for a group of unmanned aerial vehicles (UAVs). A distributed estimation architecture fusing range, visual, and intermittent position measurements is proposed. Relative range and co-observed features are utilized to construct direct and indirect geometric constraints between UAVs, respectively. Compared with its independent counterpart, the proposed collaborative estimation scheme is more accurate and robust, while maintaining scalability and efficiency in practical deployment. To solve the intractable problem of evaluating the cross covariance between local estimators during estimation, the covariance intersection (CI) algorithm is introduced into the distributed fusion scheme, where each UAV only estimates its own pose and covariance. Observability analysis is provided to gain insights about the system's identification properties. Finally, the algorithm is applied to a practical patrolling scenario of multiple UAVs, and both numerical and software-in-the-loop (SITL) simulations are performed to illustrate the feasibility and effectiveness of the proposed scheme.