CDC · 2025

INSTRUCT MPC: A Human-LLM-in-the-Loop Framework for Context-Aware Control

Ruixiang Wu, Jiahao Ai, Tongxin Li

IEEE Conference on Decision and Control

How can human instructions improve predictive control?

A language to distribution module converts contextual instructions into disturbance predictions for MPC. The framework closes the loop between instructions, predictions, and control, with a regret analysis for linear dynamics under the stated training assumptions.

Research themes

Connections

From language context to constrained actions

INSTRUCT MPC translates human instructions into disturbance predictions. World model certification and VigilMPC study complementary ways of deciding which learned models can support reliable planning and safe updates.

Context becomes an energy system input

OpenCEM motivates shared contextual energy benchmarks. INSTRUCT MPC uses instructions for control, while counterfactual load forecasting represents news events as interventions on electricity demand.

Open questions

Certified contextual control

Can contextual and world model updates be certified before they affect a physical system?

Explore open directions

Cite this paper

Ruixiang Wu, Jiahao Ai, Tongxin Li. INSTRUCT MPC: A Human-LLM-in-the-Loop Framework for Context-Aware Control. IEEE Conference on Decision and Control, 2025. https://doi.org/10.1109/CDC57313.2025.11312156

BibTeX
@inproceedings{tongxin-instruct-mpc,
  title = {{INSTRUCT MPC: A Human-LLM-in-the-Loop Framework for Context-Aware Control}},
  author = {Ruixiang Wu and Jiahao Ai and Tongxin Li},
  year = {2025},
  booktitle = {IEEE Conference on Decision and Control},
  url = {https://doi.org/10.1109/CDC57313.2025.11312156},
  doi = {10.1109/CDC57313.2025.11312156}
}
2026Applied Energy

Counterfactual load forecasting with LLM-structured events and representation learning

Yujie Chen, Yifei Gao, Runyao Yu, Yuhe Wu, Guangyu Wang, Yue Chen, Tongxin Li

NACF turns news into structured treatments and estimates load trajectories under alternative event conditions. Reweighting and representation balancing address observed confounding. Experiments examine factual accuracy and interpretable demand perturbations without claiming that unobserved counterfactual outcomes can be directly validated.

How might electricity demand change under a different news event? AI for energy electric vehicle charging demand response renewable energy load forecasting decarbonization large language models LLM agents contextual control world models reinforcement learning dueling bandits
2026IEEE Control Systems Letters

VigilMPC: Certified Online Tube Model Predictive Control for Safe Real-Time Neuromodulation

Yu Mao, Zhichao Liang, Junxiang Zhang, Quanying Liu, Tongxin Li

VigilMPC accepts a predictor update only when a backup plan remains feasible under tightened constraints. Recursive feasibility and constraint satisfaction follow under the tube MPC assumptions. Evaluation uses simulated neural dynamics and a switching benchmark, rather than clinical deployment.

Can a predictive controller safely update its learned dynamics online? model predictive control MPC stability recursive feasibility distributed control system level synthesis learning augmented algorithms algorithms with predictions competitive analysis robustness consistency online optimization
2026ICML

World Models in Pieces: Structural Certification for General Agents

Yikai Lu, Yifei Wu, Xinyu Lu, Tongxin Li

Structural certification links performance on compositional goals to local guarantees on world model transitions. The results identify reliable pieces of a model without requiring universal competence, and establish limits on the precision of such guarantees.

Which parts of an agent's world model can support reliable planning? large language models LLM agents contextual control world models reinforcement learning dueling bandits information theory graph learning sample complexity compressed sensing graph neural networks Riemannian geometry
2025ACM e-Energy demo

Open In-Context Energy Management Platform

Yikai Lu, Tinko Sebastian Bartels, Ruixiang Wu, Fanzeng Xia, Xudong Wang, Yifei Wu, Haoxiang Yang, Tongxin Li

OpenCEM presents a platform design connecting energy time series, events, human context, and simulation. An on campus solar and battery installation motivates evaluation of context sensitive control. The paper describes planned data and API capabilities.

What would a shared benchmark for contextual energy management provide? large language models LLM agents contextual control world models reinforcement learning dueling bandits AI for energy electric vehicle charging demand response renewable energy load forecasting decarbonization model predictive control MPC stability recursive feasibility distributed control system level synthesis