Safety Assessment Approach to UAVs Based on Profust Safety Index and HIL Simulation

The intelligence level of unmanned aerial vehicles (UAVs) is continuously improving, which puts forward higher requirements for reliability and safety. However, the high cost and low efficiency of experiments and the lack of failure data and safety index are important factors hindering the improvement of UAVs' safety. This article aims to design a safety assessment approach based on a proposed profust safety index and autopilot hardware-in-the-loop (HIL) simulation to provide a new solution. First, the simulation data of UAVs under fault or abnormal conditions are generated from an HIL simulation platform with high-fidelity dynamic models and 3-D visualization models. Then, the profust safety index based on the profust reliability theory is proposed to quantitatively describe the safety status of UAVs, and this article presents a novo and simple approach to calculate the accident probability. The effectiveness of the proposed approach is verified by the comparison of simulation and experiments. Finally, the impact of different failure modes on the safety of UAVs is analyzed through a large number of fault simulations of different components, and a set of simulation datasets along with the HIL platform is released to provide a reference for the safety and health assessment and fault diagnosis of UAVs.