TL;DR
This project implements AC4MPC, in which a reinforcement-learning agent is trained on the exact cost a model predictive controller optimizes, and its learned value function (critic) is embedded in the MPC as a terminal cost.
This work presents a GPU-parallelized residual architecture that tightly integrates MPC and RL by blending their outputs at the torque-control level, achieving higher sample efficiency, converges to greater asymptotic rewards, expands the range of trackable velocity commands, and enables zero-shot adaptation to unseen...
Group Relative Return Policy Optimization (GR2PO), a critic-free reinforcement learning framework for continuous robot control, is proposed and significantly outperforms critic-free baselines that use immediate rewards and performs competitively against state-of-the-art actor-critic methods.
A training method for HIL online reinforcement learning for real robots that automatically switches between learning from interventions and on-policy self-improvement, reducing the policy--target-sample gap that otherwise induces execution-time distribution shift.
This work formalizes the hybrid LLM-planner and RL-controller architecture as a Goal-Augmented Markov Decision Process and shows that when the LLM per-state progress score is used as a bounded potential function, the resulting shaping term preserves the optimal policy set even when the LLM scores are inaccurate.
This work proposes Single-rollout Autoregressive Policy Optimization (SAPO), a low-memory and compute-efficient framework in which the policy and value functions share a single autoregressive backbone, and introduces a trajectory-level generalized advantage estimator that combines lambda-returns with batch normalizatio...
A novel composite-gradient learning (CGL) method is proposed that integrates the MPC controller into the learning process by representing the DRL and MPC control inputs as a joint action and accounting for their interaction when updating the DRL agent during training.
We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.
