Robust Value Iteration for Continuous Control Tasks

05/25/2021
by   Michael Lutter, et al.
2

When transferring a control policy from simulation to a physical system, the policy needs to be robust to variations in the dynamics to perform well. Commonly, the optimal policy overfits to the approximate model and the corresponding state-distribution, often resulting in failure to trasnfer underlying distributional shifts. In this paper, we present Robust Fitted Value Iteration, which uses dynamic programming to compute the optimal value function on the compact state domain and incorporates adversarial perturbations of the system dynamics. The adversarial perturbations encourage a optimal policy that is robust to changes in the dynamics. Utilizing the continuous-time perspective of reinforcement learning, we derive the optimal perturbations for the states, actions, observations and model parameters in closed-form. Notably, the resulting algorithm does not require discretization of states or actions. Therefore, the optimal adversarial perturbations can be efficiently incorporated in the min-max value function update. We apply the resulting algorithm to the physical Furuta pendulum and cartpole. By changing the masses of the systems we evaluate the quantitative and qualitative performance across different model parameters. We show that robust value iteration is more robust compared to deep reinforcement learning algorithm and the non-robust version of the algorithm. Videos of the experiments are shown at https://sites.google.com/view/rfvi

READ FULL TEXT

page 2

page 6

page 8

page 16

page 17

research
10/05/2021

Continuous-Time Fitted Value Iteration for Robust Policies

Solving the Hamilton-Jacobi-Bellman equation is important in many domain...
research
02/21/2020

Estimating Q(s,s') with Deep Deterministic Dynamics Gradients

In this paper, we introduce a novel form of value function, Q(s, s'), th...
research
05/10/2021

Value Iteration in Continuous Actions, States and Time

Classical value iteration approaches are not applicable to environments ...
research
02/27/2023

Distributional Method for Risk Averse Reinforcement Learning

We introduce a distributional method for learning the optimal policy in ...
research
06/21/2022

Robust Deep Reinforcement Learning through Bootstrapped Opportunistic Curriculum

Despite considerable advances in deep reinforcement learning, it has bee...
research
06/10/2020

Distributional Robust Batch Contextual Bandits

Policy learning using historical observational data is an important prob...
research
07/14/2020

Ternary Policy Iteration Algorithm for Nonlinear Robust Control

The uncertainties in plant dynamics remain a challenge for nonlinear con...

Please sign up or login with your details

Forgot password? Click here to reset