Catch Planner: Catching High-Speed Targets in the Flight

TL;DR

Catch Planner is a novel integrated planning and decision-making system that enables quadrotors to catch high-speed flying targets effectively by combining deep reinforcement learning and trajectory optimization, demonstrated through real and simulated experiments.

Contribution

The paper introduces Catch Planner, a new framework combining learning and planning for high-speed target catching, with a flexible constraint transcription method and real-time onboard implementation.

Findings

Robustness demonstrated in diverse real and simulated scenarios.

Achieves high success rate in catching high-speed targets.

Operates at 100Hz on onboard hardware.

Abstract

Catching high-speed targets in the flight is a complex and typical highly dynamic task. In this paper, we propose Catch Planner, a planning-with-decision scheme for catching. For sequential decision making, we propose a policy search method based on deep reinforcement learning. In order to make catching adaptive and flexible, we propose a trajectory optimization method to jointly optimize the highly coupled catching time and terminal state while considering the dynamic feasibility and safety. We also propose a flexible constraint transcription method to catch targets at any reasonable attitude and terminal position bias. The proposed Catch Planner provides a new paradigm for the combination of learning and planning and is integrated on the quadrotor designed by ourselves, which runs at 100hz on the onboard computer. Extensive experiments are carried out in real and simulated scenes to verify the robustness of the proposed method and its expansibility when facing a variety of high-speed flying targets.