Research/ SPACE AI & AUTONOMY

On Deep Recurrent Reinforcement Learning for Active Visual Tracking of Space Noncooperative Objects

Dong Zhou, Guanghui Sun, Zhao Zhang, Ligang Wu

RA-L 2022/Dec 29, 2022/ 30 citations

Space × Robotics × AI

Technique

Learning-based MethodsOpenAlex topic: Reinforcement Learning in Robotics · Rendezvous, Proximity Ops & DockingOpenAlex keyword: autonomous rendezvous

Indexed as

arXivcs.ROOpenAlexReinforcement Learning in Robotics · Space Satellite Systems and Control · Spacecraft Dynamics and Control

ar5iv HTML PDF arXiv:2212.14304 DOI Code Semantic Scholar OpenAlex

Active tracking of space noncooperative object that merely relies on vision camera is greatly significant for autonomous rendezvous and debris removal. Considering its Partial Observable Markov Decision Process (POMDP) property, this paper proposes a novel tracker based on deep recurrent reinforcement learning, named as RAMAVT which drives the chasing spacecraft to follow arbitrary space noncooperative object with high-frequency and near-optimal velocity control commands. To further improve the active tracking performance, we introduce Multi-Head Attention (MHA) module and Squeeze-and-Excitation (SE) layer into RAMAVT, which remarkably improve the representative ability of neural network with almost no extra computational cost. Extensive experiments and ablation study implemented on SNCOAT benchmark show the effectiveness and robustness of our method compared with other state-of-the-art algorithm. The source codes are available on https://github.com/Dongzhou-1996/RAMAVT.

On Deep Recurrent Reinforcement Learning for Active Visual Tracking of Space Noncooperative Objects

Related papers

Space Non-cooperative Object Active Tracking with Deep Reinforcement Learning

Deep Reinforcement Learning for Spacecraft Proximity Operations Guidance

Sample-Efficient and Robust Visual Reinforcement Learning for Autonomous Spacecraft Proximity Operations

Spacecraft Adaptive Deep Reinforcement Learning Guidance with Input State Uncertainties in Relative Motion Scenario

Spacecraft Proximity Maneuvering and Rendezvous With Collision Avoidance Based on Reinforcement Learning

Recurrent Vision Transformers for Event-Based Spacecraft Pose Estimation Under Slow Relative Motions

Revisiting Space Mission Planning: A Reinforcement Learning-Guided Approach for Multi-Debris Rendezvous

Reinforcement Learning-Based Trajectory Planning for Tracking and Rendezvous of Small Spacecraft

Action Chunking with Transformers for Image-Based Spacecraft Guidance and Control

Image-Based Meta-Reinforcement Learning for Autonomous Guidance of an Asteroid Impactor

More from RA-L

Exploring Event Camera-based Odometry for Planetary Robots

Challenges of SLAM in extremely unstructured environments: the DLR Planetary Stereo, Solid-State LiDAR, Inertial Dataset

MLNav: Learning to Safely Navigate on Martian Terrains

An Approach to Science and Risk-Aware Planetary Rover Exploration

On Deep Recurrent Reinforcement Learning for Active Visual Tracking of Space Noncooperative Objects

Related papers

Space Non-cooperative Object Active Tracking with Deep Reinforcement Learning

Deep Reinforcement Learning for Spacecraft Proximity Operations Guidance

Sample-Efficient and Robust Visual Reinforcement Learning for Autonomous Spacecraft Proximity Operations

Spacecraft Adaptive Deep Reinforcement Learning Guidance with Input State Uncertainties in Relative Motion Scenario

Spacecraft Proximity Maneuvering and Rendezvous With Collision Avoidance Based on Reinforcement Learning

Recurrent Vision Transformers for Event-Based Spacecraft Pose Estimation Under Slow Relative Motions

Revisiting Space Mission Planning: A Reinforcement Learning-Guided Approach for Multi-Debris Rendezvous

Reinforcement Learning-Based Trajectory Planning for Tracking and Rendezvous of Small Spacecraft

Action Chunking with Transformers for Image-Based Spacecraft Guidance and Control

Image-Based Meta-Reinforcement Learning for Autonomous Guidance of an Asteroid Impactor

More from RA-L

Exploring Event Camera-based Odometry for Planetary Robots

Challenges of SLAM in extremely unstructured environments: the DLR Planetary Stereo, Solid-State LiDAR, Inertial Dataset

MLNav: Learning to Safely Navigate on Martian Terrains

An Approach to Science and Risk-Aware Planetary Rover Exploration