ChangHun Ji, Youn-Hee Han,

Journal of Advanced Technology Research, Vol. 8, No. 1, pp. 15-19, Jun. 2023
10.11111/JATR.2023.8.1.015, Full Text:
Keywords: Drone Autonomous Flight, EGO-Swarm, Hierarchical Deep Reinforcement Learning
Abstract

The EGO-Swarm system has recently gained attention for its successful generation of smooth drone flight paths in complex real-world environments. EGO-Swarm utilizes various parameter values for smooth path generation, and among these parameters, the maximum speed and maximum acceleration of the drone play a crucial role in improving flight performance. However, in EGO-Swarm, the initially set parameter values remain fixed during path creation, despite the fact that the optimal values for maximum speed and maximum acceleration may change in real-time dynamic environments. Therefore, it is desirable to dynamically adjust these values according to the changing real-time environment. In this paper, we propose a novel algorithm that dynamically sets the maximum speed and maximum acceleration using hierarchical deep reinforcement learning in response to real-time environmental changes. To validate the effectiveness of the proposed method, a comparative experiment between the existing EGO-Swarm algorithm and the proposed algorithm is conducted in a ROS simulation. The experimental results demonstrate that the proposed algorithm outperforms the existing EGOSwarm algorithm in terms of average speed improvement and path length.

Statistics
Cite this article