Risk-Based UAV Flight Path Optimization in Accordance with SORA

Abstract

With the EU regulation of drone operations varying based on the specific type of drone, path planning can be done to consider risk mitigation. This paper proposes a transition-based rapidly exploring random tree star (T-RRT*) path planning algorithm for fixed-wing drone operations over rural areas. Risk is decoded in the cost function, which mainly considers population density and special infrastructure types. It was found that the algorithm is capable of finding paths that minimize exposure of the general population and infrastructure. However, path and computational inefficiencies were found. Usage of another data structure or algorithm might improve performance.

Keywords

• optimal path planning
• complex cost-space planning
• sampling-based path planning
• T-RRT*
• rural drone operations
• (EU) VO 2019/947