Cooperative differential relative localization method for UAV formation based on MSIF

DOI:	10.11809/bqzbgcxb2025.01.032
Keywords:	unmanned aerial vehicle formation; multi sensor information fusion; relative localization; cooperative localization; datalink range
Abstract:	To address the problem of relative localization for UAVs formations under GNSS denied conditions, a cooperative differential relative localization method is proposed, based on multi sensor information fusion (MSIF) and datalink range. Relative localization can be achieved by utilizing range and relative velocity measurements in a Kalman filtering paradigm. But when formation members move in a same velocity the observability conditions cannot be met, and the filtering method fails. To address the challenge mentioned above, a cooperative differential relative localization algorithm is proposed. An INS/barometer integrated navigation system is implemented to ensure the height accuracy. Then the localization problem is transformed into 2 D plane, the range and IMU information are utilized to correct the horizontal position errors of followers. By properly arranging the distance between leader and follower, the remaining error of followers are approximately identical. Finally, the remaining common error is eliminated in a difference calculation and the accurate relative localization between two followers is obtained. The Monte Carlo simulation demonstrates that the method can accurately estimate relative position in an UAVs formation and keep the accuracy when formation members move in a same velocity. While the condition of leader follower distance is satisfied, the estimation accuracy is determined by range accuracy and free from the limit of IMU accuracy. The relative positioning accuracy is around 2.3, 5.8, and 11.7 m for ranging accuracies of 1, 5, and 10 m, respectively. The method is feasible for short or medium endurance UAVs formations.
Issue:	Vol. 46 No. 1 (2025)
Published:	2025-01-31
PDF