UAV Optimal Obstacle Avoidance while Respecting Target Arrival Specifications

Abstract

AbstractEffective target tracking and obstacle avoidance strategies are essential to the success of unmanned aerial vehicle (UAV) missions. This paper describes a decentralized model-based predictive control to calculate the optimal UAV trajectory that will lead each UAV to the interception of a target at a desired time and with desired speed, heading and flight path angle, while avoiding dynamic ellipsoidal obstacles detected en route. Obstacle avoidance is based on the minimization of the UAV collision probability with all known obstacles on a future horizon, while ensuring that the collision probability with any given obstacle at each prediction step does not surpass a preset threshold. Cooperation between UAVs is possible by exchanging information about the obstacles they detect. To respect the arrival specifications, a virtual target is created and tracked. The virtual target moves at the desired speed and along a straight line that is correctly oriented to intercept the real target position at the desired arrival time. Arrival specifications are respected if tracking of the virtual target by the UAV is successful. Simulations are presented to demonstrate the effectiveness of the proposed approach. They also illustrate that cooperation can improve the UAV performance.