Navigation information augmented artificial potential field algorithm for collision avoidance in UAV formation flight

Abstract

In recent years, multiple Unmanned Aerial Vehicle (UAV) formation flight has attracted worldwide research interest, for its potential benefits of scalability and flexibility. In complex urban environments, the successful operation of those UAVs requires the system to provide certain safety level. As one of the key requirements, collision avoidance improves the system's ability to accommodate operational environment variations, and to perform multiple tasks. To achieve this, artificial potential field (APF) has been recognized as one of the most suitable methods along with drone control. Although there has been substantial relevant work on the APF for single UAV in static environment, more efforts are desired to address formation maneuvers in complex environments such as urban. Most importantly, the traditional APF algorithms do not account for random errors in navigation solutions, which can bring potential risk to the UAV system. In response, this paper proposes a new APF algorithm that employs navigation information in complex urban environments, and the goal is to realize UAV formation collision avoidance. By augmenting the APF algorithm with UAV navigation information, the potential risk caused by navigation uncertainty can be mitigated, especially in the Global Navigation Satellite System (GNSS) challenged environment. The principle of the new APF approach is adaptively estimating the parameters of potential field force function, using the variance of navigation information and user-defined confidence probability. This new approach is applied in the synchronized UAV formation collision avoidance control. As a result, the UAVs can achieve fast position and attitude adjustment with high safety confidence. To verify the algorithm, quadrotors with emulated GNSS receivers are used to generate observation data. These data are incorporated into a complex urban environment simulation, where multiple sets of virtual obstacles are injected. Results show that the proposed method can achieve safe and efficient collision avoidance for cooperative formation flight in urban GNSS challenged environment.

A Review of Consensus-based Multi-agent UAV Implementations

“…Several applications employing multi-rotor UAVs as delivery services, bridge inspection, and traffic monitoring are deployed in urban environments, where an adequate safety level must be maintained [13]. A crucial requirement in such conditions is the capability of generating a formation while avoiding collisions between members of the swarm and crashes with external obstacles [14].…”

Motion Regulation Solutions to Holding and Moving an Object for Single-Leader–Dual-Follower Teleoperation

“…To improve the ability of motion regulation of a robot, the artificial potential field method (APFM) has been verified to be a practical means. Its use is found to cover a large number of application scenarios, which are mostly about mobile robots' or unmanned aerial vehicles' (UAVs') [16] path planing. There are few works about APFM for manipulators' motion regulation [17], [18].…”

Learning based multi-obstacle avoidance of unmanned aerial vehicles with a novel reward

“…This method involves converting the influence of the target and obstacle on the UAV' s movement into an artificial potential field. By utilizing the gravitational and repulsive forces between the UAV and the obstacle, the researchers were able to effectively control the UAV' s movement towards the target point, following the negative gradient of the potential field [8] . The aforementioned methods pertain to the category of classical collision avoidance techniques.…”