A Real-Time Rerouting Method for Drone Flights Under Uncertain Flight Time

Abstract

This paper proposes a method for real-time rerouting drone flights under uncertain flight times. The battery runtime that remains of a drone in real-time may be insufficient to complete its flight mission. This may be due to external factors, such as unexpected severe weather or obstacles that move into the drone’s flight path. Under unexpected situations, such as these, the drone cannot safely return to its depot, as planned. To ensure that the drone makes a safe return and that the flight mission is a success, there must be a real-time rerouting process for a drone’s flight path in response to unforeseeable circumstances. Hence, this paper proposes a real-time rerouting process consisting of two optimization models that generate an optimal alternative flight path for a drone that has insufficient remaining battery runtime. The first model is used to find an optimal flight path to visit all remaining target waypoints. If the first model fails to obtain a feasible solution, the second model is implemented to find an optimal flight path to minimize the number of unvisited waypoints. To confine the total flight (travel) time to the insufficient battery runtime, both models include the constraint associated with uncertain flight (travel) times between waypoints. To capture this uncertainty, this paper proposes a chance constrained programming (CCP) approach under the assumption of a known mean, variance, and flight time intervals. Numerical examples show how the proposed rerouting process works, and the CCP method results in more conservative solutions as compared to the deterministic approach.