ROAILGSYMar 13, 2022

Adaptive Model Predictive Control by Learning Classifiers

arXiv:2203.06783v22 citationsh-index: 21
Originality Incremental advance
AI Analysis

This work addresses adaptive control for robotics tasks with model uncertainty, representing an incremental improvement by integrating existing methods into a new framework.

The paper tackles the problem of adjusting control parameters in stochastic model predictive control under model uncertainty and heteroscedastic noise by proposing an adaptive MPC variant that uses Bayesian optimization reformulated via density ratio estimation, approximated by learning a classifier, and demonstrates it on control problems and robotics manipulation tasks.

Stochastic model predictive control has been a successful and robust control framework for many robotics tasks where the system dynamics model is slightly inaccurate or in the presence of environment disturbances. Despite the successes, it is still unclear how to best adjust control parameters to the current task in the presence of model parameter uncertainty and heteroscedastic noise. In this paper, we propose an adaptive MPC variant that automatically estimates control and model parameters by leveraging ideas from Bayesian optimisation (BO) and the classical expected improvement acquisition function. We leverage recent results showing that BO can be reformulated via density ratio estimation, which can be efficiently approximated by simply learning a classifier. This is then integrated into a model predictive path integral control framework yielding robust controllers for a variety of challenging robotics tasks. We demonstrate the approach on classical control problems under model uncertainty and robotics manipulation tasks.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes