A Real-Time Collision Avoidance Strategy in Dynamic Airspace Based on Dynamic Artificial Potential Field Algorithm

Abstract

The key to the integration of unmanned aircrafts in the national airspace is to prevent them from colliding with the other traffics in the airspace. However, it is a great challenge to generate a safe, stable and robust collision-free path for unmanned aircraft system (UAS), a.k.a. unmanned aircraft vehicle (UAV), in real time, if the airspace is highly dynamics and heterogenous (i.e. thronged with aircrafts with different motion states). Based on the dynamic artificial potential field (DAPF) algorithm, this paper provides advisories on how to generate a real-time reactive collision-free path for unmanned aircraft vehicles flying in a dynamic airspace, aiming to ensure the flight safety and minimize the impact on surrounding traffic. Firstly, the safety distance was defined as a variable threshold, which scaled adaptively according to the relative motion states of the surrounding obstacles and the performance of the own UAV. Moreover, the forces of the potential field were improved, such that their magnitudes could be adjusted automatically according to the threat levels of the surrounding obstacles. The threat level of an obstacle depends on the relative position, speed and flight trend between the UAV and the obstacle. In addition, the repulsive force along the relative position of the traditional artificial potential field (APF) was retained, and a steering force was added to change the flight direction of the UAV, aiming to speed up the collision avoidance. Furthermore, the attractive force was modified to help the UAV return to the planned path quickly and stably. After that, the capacity were determined to ensure the feasible and practical path planning for the UAV. Finally, the proposed UAV path-planning method was proved effective, safe, stable and adaptive through simulations.