Collected molecules will appear here. Add from search or explore.
Uses Data-enabled Predictive Control (DeePC) as a real-time compensator to bridge the gap between simulation-trained RL policies and physical robotic dynamics without retraining the policy.
stars
0
forks
0
This is a specialized research project combining control theory (DeePC) with modern RL. While the approach of using an external controller to regulate the sim2real gap is academically interesting, the repository has zero stars or forks and is very recent, indicating it's likely a personal research codebase or a companion to a specific paper. It lacks the community or infrastructure to be defensive.
TECH STACK
INTEGRATION
reference_implementation
READINESS