Diffusion-Based Forecasting for Uncertainty-Aware Model Predictive Control

2025-03-19

AI Summary

To address uncertainty-aware prediction and decision-making in partially observable stochastic systems, this paper proposes Diffusion-Informed MPC—a novel framework that integrates diffusion models into model predictive control (MPC) for the first time, enabling end-to-end uncertainty propagation modeling. The method supports probabilistic time-series forecasting under non-Gaussian and arbitrarily structured noise, combining Bayesian trajectory sampling with uncertainty-constrained receding-horizon optimization. Evaluated on a battery energy storage arbitrage task in the New York State day-ahead electricity market, it achieves 17.3% and 22.8% higher revenue than statistical-forecasting-based MPC and model-free reinforcement learning baselines, respectively, while significantly improving robustness and risk management. Its core contribution lies in overcoming traditional MPC’s reliance on deterministic or Gaussian assumptions, establishing a scalable, probabilistic closed-loop decision-making paradigm.

Technology Category

• Reasoning under Uncertainty: Stochastic Optimization
• Multiagent Systems: Multiagent Systems under Uncertainty
• Planning, Routing, and Scheduling: Scheduling under Uncertainty

Application Category

• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• User Modeling, Personalization and Recommendation: Psychology-informed user models and recommender systems

Abstract

We propose Diffusion-Informed Model Predictive Control (D-I MPC), a generic framework for uncertainty-aware prediction and decision-making in partially observable stochastic systems by integrating diffusion-based time series forecasting models in Model Predictive Control algorithms. In our approach, a diffusion-based time series forecasting model is used to probabilistically estimate the evolution of the system's stochastic components. These forecasts are then incorporated into MPC algorithms to estimate future trajectories and optimize action selection under the uncertainty of the future. We evaluate the framework on the task of energy arbitrage, where a Battery Energy Storage System participates in the day-ahead electricity market of the New York state. Experimental results indicate that our model-based approach with a diffusion-based forecaster significantly outperforms both implementations with classical forecasting methods and model-free reinforcement learning baselines.

Problem

• Uncertainty-aware prediction in stochastic systems
• Integration of diffusion models in MPC algorithms
• Optimizing energy arbitrage in electricity markets

Innovation

• Integrates diffusion models in MPC for uncertainty-aware control.
• Uses probabilistic forecasting for system evolution estimation.
• Optimizes actions under future uncertainty in energy markets.