Abstract:

In order to ensure the flight safety of multi-rotor UAVs in low-altitude airspace, a conflict detection model and conflict resolution strategy for UAV low-altitude flights were proposed. According to the current Automatic Dependent Surveillance-Broadcast (ADS-B) data of the UAV, the interactive multi-model volumetric Kalman filtering algorithm (IMM-CKF) was used to predict the trajectory of the next stage, and the potential conflict objects were screened. According to the characteristics of the flight protection area, the AABB bounding boxes of the two aircrafts are constructed, and the three-dimensional collision detection algorithm was used to judge whether the two aircraft conflict, and SOSA flight conflict resolution strategy was proposed. Finally, four common conflict scenarios were constructed. The proposed algorithm was used to detect flight conflict. The SOSA combined resolution strategy and the traditional resolution strategy, such as the height adjustment method, the direction adjustment method and the speed adjustment method were used to resolve the flight conflict according to the situation. The results show that the flight conflict detection method based on the AABB bounding box can quickly detect conflicts within 0.02 s, and the use of SOSA combined resolution strategy can reduce the increase of the flying distance to less than 5% of the direct flight distance compared with the traditional resolution strategy. The conflict detection method is robust in real time, the complexity of the resolution strategy operation is low, and the payment distance is small during the conflict resolution process.

Key words: geometric relations, unmanned aerial vehicle(UAV), low altitude flight, conflict detection, resolution strategy, same orientation same angle(SOSA)