Research on experimental evaluation model of unmanned aerial vehicle autonomy based on digital twin

Abstract

Abstract The current focus of unmanned aerial vehicle (UAV) research lies in the scientific evaluation of UAV autonomy and the enhancement of belief in UAVs. Researchers are required to conduct numerous tests to assess the ability of UAVs to perform missions in complex environments. However, conducting real-world tests for UAVs costs a lot. Therefore, digital twin virtual simulation serves as a viable alternative to actual flight tests. Firstly, this paper establishes a digital twin framework that enables comprehensive perception and integration within a virtual environment, proposing a UAV autonomy experimental evaluation model. Additionally, an effective method for evaluating UAV autonomy is constructed, and test tasks targeting autonomous capability assessment are designed by using simulation software and verified through system examples. Finally, based on the simulation results, the autonomous capability of the UAV twin is evaluated and utilized as a reference for assessing the autonomy of physical UAVs, thereby providing support for scientific evaluation.