Minimizing Relative Position and Attitude Error in UAV Swarms via General Form Consensus

Xander Jones, Eryn Jaramillo, Daniel Kimball-Garrett Mills, David L. Olson, Stephen Bruder, and Aly El-Osery

Peer Reviewed

Abstract:	This paper explores the use of general form consensus (GFC) in cooperative navigation among unmanned aerial vehicles (UAVs), drawing an analogy with marching band performers who maintain relative positioning without absolute position knowledge, relying only on peripheral vision. In UAV swarms, managing the absolute and relative position, velocity, and attitude (PVA) error drift is crucial. By implementing GFC, this research aims to reduce the cost, size, weight, and power (C-SWAP) requirements for each UAV and decrease dependence on high-quality navigation sensors, thus lowering the barrier for UAV swarm applications in academic and commercial fields. The paper focuses on improving relative PVA performance, while also acknowledging the potential inclusion of GPS receivers in some UAVs for enhanced long-term absolute positioning. Cooperative navigation through GFC demonstrates notable advantages over individual dead-reckoning methods, particularly in reducing relative position and attitude errors. Monte Carlo analysis is used to validate the GFC approach, examining navigation performance benefits in relation to the number of swarm members and the quality of their onboard navigation sensors. Ultimately, GFC not only achieves long-term accurate navigation for the entire swarm but also allows for a reduction in individual UAVs’ C-SWAP requirements. This paper explores the possibility of using solution-level aiding in conjunction with General Form Consensus to establish bounds on both relative and absolute positional and orientation error using only relative sensors. Preliminary results show an improvement in the scale of errors in the relative and absolute domains for both positioning and orientation. The ability of GFC to establish bounds on error or the drift of error over time is observed to be impacted by the velocity random walk (VRW) of the inertial measurement unit (IMU) used. Furthermore, the performance of the filter is observed to improve as more agents are included in consensus.
Published in:	Proceedings of the ION 2024 Pacific PNT Meeting April 15 - 18, 2024 Hilton Waikiki Beach Honolulu, Hawaii
Pages:	649 - 663
Cite this article:	Jones, Xander, Jaramillo, Eryn, Mills, Daniel Kimball-Garrett, Olson, David L., Bruder, Stephen, El-Osery, Aly, "Minimizing Relative Position and Attitude Error in UAV Swarms via General Form Consensus," Proceedings of the ION 2024 Pacific PNT Meeting, Honolulu, Hawaii, April 2024, pp. 649-663. https://doi.org/10.33012/2024.19623
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In