A fault risk learning-based safety control method for UAVs

Key Points

• The aim is to develop a safety control method for UAVs that adapts to fault risks affecting their movement.
• Quantified fault risk using conditional value-at-risk (CVaR) based on fixed-time observations.
• Collected experimental data to analyze uncertainty in UAV positioning under varying fault risk levels.
• Utilized a sliding time window and lightweight backpropagation neural network to learn the impact of fault risk.
• Employed an adaptive risk inclination control strategy for smooth adjustments in response to fault risk.
• The adaptive control strategy improved response speed and trajectory tracking accuracy at various fault levels.
• Compared to traditional sliding mode control, the new method maintained performance with lighter computational demands.