Datamodel-Based Data Selection for Nonlinear Data-Enabled Predictive Control

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Li, Jiachen, Li, Shihao, Xu, Jiamin, Bakshi, Soovadeep, Chen, Dongmei
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918407246446592
author Li, Jiachen
Li, Shihao
Xu, Jiamin
Bakshi, Soovadeep
Chen, Dongmei
author_facet Li, Jiachen
Li, Shihao
Xu, Jiamin
Bakshi, Soovadeep
Chen, Dongmei
contents Data-Enabled Predictive Control (DeePC) has emerged as a powerful framework for controlling unknown systems directly from input-output data. For nonlinear systems, recent work has proposed selecting relevant subsets of data columns based on geometric proximity to the current operating point. However, such proximity-based selection ignores the control objective: different reference trajectories may benefit from different data even at the same operating point. In this paper, we propose a datamodel-based approach that learns a context-dependent influence function mapping the current initial trajectory and reference trajectory to column importance scores. Adapting the linear datamodel framework from machine learning, we model closed-loop cost as a linear function of column inclusion indicators, with coefficients that depend on the control context. Training on closed-loop simulations, our method captures which data columns actually improve tracking performance for specific control tasks. Experimental results demonstrate that task-aware selection substantially outperforms geometry-based heuristics, particularly when using small data subsets.
format Preprint
id arxiv_https___arxiv_org_abs_2512_00276
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Datamodel-Based Data Selection for Nonlinear Data-Enabled Predictive Control
Li, Jiachen
Li, Shihao
Xu, Jiamin
Bakshi, Soovadeep
Chen, Dongmei
Systems and Control
Data-Enabled Predictive Control (DeePC) has emerged as a powerful framework for controlling unknown systems directly from input-output data. For nonlinear systems, recent work has proposed selecting relevant subsets of data columns based on geometric proximity to the current operating point. However, such proximity-based selection ignores the control objective: different reference trajectories may benefit from different data even at the same operating point. In this paper, we propose a datamodel-based approach that learns a context-dependent influence function mapping the current initial trajectory and reference trajectory to column importance scores. Adapting the linear datamodel framework from machine learning, we model closed-loop cost as a linear function of column inclusion indicators, with coefficients that depend on the control context. Training on closed-loop simulations, our method captures which data columns actually improve tracking performance for specific control tasks. Experimental results demonstrate that task-aware selection substantially outperforms geometry-based heuristics, particularly when using small data subsets.
title Datamodel-Based Data Selection for Nonlinear Data-Enabled Predictive Control
topic Systems and Control
url https://arxiv.org/abs/2512.00276