DOMA: Deep Smooth Trajectory Generation Learning for Real-Time UAV Motion Planning

Authors

  • Jin Yu SADRI Institute
  • Haiyin Piao Northwestern Polytechnical University
  • Yaqing Hou Dalian University of Technology
  • Li Mo Beijing Institute of Technology
  • Xin Yang Dalian University of Technology
  • Deyun Zhou Northwestern Polytechnical University

Keywords:

Motion Planning, UAV, Deep Reinforcement Learning, Smooth

Abstract

In this paper, we present a Deep Reinforcement Learning (DRL) based real-time smooth UAV motion planning method for solving catastrophic flight trajectory oscillation issues. By formalizing the original problem as a linear mixture of dual-objective optimization, a novel Deep smOoth Motion plAnning (DOMA) algorithm is proposed, which adopts an alternative layer-by-layer gradient descending optimization approach with the major gradient and the DOMA gradient applied separately. Afterward, the mix weight coefficient between the two objectives is also optimized adaptively. Experimental result reveals that the proposed DOMA algorithm outperforms baseline DRL-based UAV motion planning algorithms in terms of both learning efficiency and flight motion smoothness. Furthermore, the UAV safety issue induced by trajectory oscillation is also addressed.

Downloads

Published

2022-06-13

How to Cite

Yu, J., Piao, H., Hou, Y., Mo, L., Yang, X., & Zhou, D. (2022). DOMA: Deep Smooth Trajectory Generation Learning for Real-Time UAV Motion Planning. Proceedings of the International Conference on Automated Planning and Scheduling, 32(1), 662-666. Retrieved from https://ojs.aaai.org/index.php/ICAPS/article/view/19855