Adversarially Regularized Policy Learning Guided by Trajectory Optimization
This work addresses the challenge of learning smooth and robust control policies for robot manipulation, representing an incremental improvement over existing methods by integrating adversarial regularization.
The paper tackles the problem of overcomplex and non-smooth neural control policies in robot systems, which cause poor generalization, by proposing VERONICA, an adversarially regularized policy learning method guided by trajectory optimization. The result shows improved sample efficiency and enhanced robustness against disturbances like sensor noise and environmental uncertainty in robot manipulation experiments.
Recent advancement in combining trajectory optimization with function approximation (especially neural networks) shows promise in learning complex control policies for diverse tasks in robot systems. Despite their great flexibility, the large neural networks for parameterizing control policies impose significant challenges. The learned neural control policies are often overcomplex and non-smooth, which can easily cause unexpected or diverging robot motions. Therefore, they often yield poor generalization performance in practice. To address this issue, we propose adVErsarially Regularized pOlicy learNIng guided by trajeCtory optimizAtion (VERONICA) for learning smooth control policies. Specifically, our proposed approach controls the smoothness (local Lipschitz continuity) of the neural control policies by stabilizing the output control with respect to the worst-case perturbation to the input state. Our experiments on robot manipulation show that our proposed approach not only improves the sample efficiency of neural policy learning but also enhances the robustness of the policy against various types of disturbances, including sensor noise, environmental uncertainty, and model mismatch.