Abstract
We introduce a simple but effective method for managing risk in zero-order trajectory optimization that involves probabilistic safety constraints and balancing of optimism in the face of epistemic uncertainty and pessimism in the face of aleatoric uncertainty of an ensemble of stochastic neural networks. Various experiments indicate that the separation of uncertainties is essential to performing well with data-driven MPC approaches in uncertain and safety-critical control environments.
Media
Experiments
Solo8-LeanOverObject
Noisy-FetchPickAndPlace
Additional Information
Links
BibTex
@inproceedings{
vlastelica2021riskaverse,
title = {Risk-Averse Zero-Order Trajectory Optimization},
author = {Marin Vlastelica and Sebastian Blaes and Cristina Pinneri and Georg Martius},
booktitle = {5th Annual Conference on Robot Learning },
year = {2021},
url = {https://openreview.net/forum?id=WqUl7sNkDre}
}
title = {Risk-Averse Zero-Order Trajectory Optimization},
author = {Marin Vlastelica and Sebastian Blaes and Cristina Pinneri and Georg Martius},
booktitle = {5th Annual Conference on Robot Learning },
year = {2021},
url = {https://openreview.net/forum?id=WqUl7sNkDre}
}