Trajectory Risk Evaluation for Autonomous Low-Flying Air Transport

The rising effort of rulemaking authorities to open the sky for the civil use of large unmanned autonomous aircraft flying beyond the line of sight created interest to find safe flight paths for these vehicles to protect the environment. One main aspect of this problem is to mitigate the situation of total loss of control over the vehicle followed by an impact somewhere on the ground resulting in possible human injuries/fatalities or infrastructure damage. The first step for such mitigation is to gather information and data about the ground to be able to avoid areas of dense population or, in general, of high damage potential. The second step, which is the focus of this paper, is the generation of flight paths with respect to a minimal risk of damage on the ground. The approach presented in this paper uses a time-discrete forward prediction method to create a normally distributed impact area for a vehicle equipped with an airframe parachute which is deployed in case of an imminent loss of control. The proposed algorithm evaluates the risk of a preplanned flight path using a newly developed impact location model based on an expandable set of nonlinear state equations and can be integrated in any existing path planning algorithm.