UAV Optimal Cooperative Target Tracking and Collision Avoidance of Moving Objects

Abstract

Effective target tracking and collision avoidance schemes are essential to the success of unmanned aerial vehicle (UAV) missions. In a dynamic environment, UAV path planning often relies on predicted obstacle and target motion. This paper presents an algorithm that predicts the trajectory of a moving object (target or obstacle) detected by a UAV. An extended Kalman filter is first employed to estimate the states of the object from its measured spatial position. The optimal object trajectory and its associated position prediction error are then calculated using the state equation defined for Kalman filtering. The proposed trajectory prediction scheme is afterward tested in a path planner which relies on decentralized cooperative predictive control to select optimal UAV trajectories as a function of the predicted target and obstacle trajectories. Simulation results are presented to demonstrate the effectiveness of the proposed approach.