Soft Actor-Critic outperforms Proximal Policy Optimization in dual-arm space robot control

A new research paper compares Proximal Policy Optimization (PPO) and Soft Actor-Critic (SAC) for controlling dual-arm space robots in cooperative motion tasks. The study found that SAC significantly outperformed PPO in reaching targets, maintaining orientation, and avoiding collisions, achieving an 80% success threshold with a median of 1.6 million environment steps compared to PPO's failure to reach this threshold within 4 million steps. While both algorithms showed degradation on unseen targets, SAC maintained most of its performance, whereas PPO's performance collapsed, highlighting SAC's robustness and PPO's sensitivity to reward shaping. AI

IMPACT Demonstrates superior performance of Soft Actor-Critic over Proximal Policy Optimization for complex robotic control tasks, potentially influencing future research in space robotics.

RANK_REASON Academic paper comparing two reinforcement learning algorithms for a specific robotics application. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• dual-arm space robots
• Nasim Fallahiarezoodar
• on-orbit servicing
• Proximal Policy Optimization
• Soft Actor-Critic

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →