Distributionally Robust Model Predictive Control for Virtual Power Plants
Distributionally Robust Model Predictive Control for Virtual Power Plants
Authors: Nikolas Recke, Mathias Hudoba de Badyn Date: 2026-05-14 Paper ID: openalex:2605.14642
Summary
This paper presents a distributionally robust model predictive control (DR-MPC) framework for Virtual Power Plant (VPP) operation under electricity price uncertainty. By integrating data-driven forecasts with time-varying Wasserstein ambiguity sets, the approach allows the VPP controller to adapt to forecast dispersion and potential distribution shifts. Evaluations on a Nordic case study demonstrate that the proposed method improves economic performance over standard MPC when the ambiguity radius is appropriately tuned, balancing robust reliability with conservativeness.
Key Contributions
- Proposes a DR-MPC framework for Virtual Power Plants (VPPs) that integrates data-driven forecasting with quantile-based uncertainty quantification.
- Introduces time-varying Wasserstein ambiguity sets to adapt control actions to forecast dispersion and distributional shifts.
- Demonstrates up to 0.8% economic performance improvement in Nordic VPP scenarios compared to standard forecast-based MPC.
Open Questions & Future Work
Key Concepts
- time-varying-wasserstein-ambiguity-sets: A mechanism for constructing robust decision-making sets that adapt to time-varying forecast uncertainty and shifts in data distributions.
Archivist Review
The paper proposes a robust control framework that dynamically adapts to forecast uncertainty. I approved the ‘time-varying Wasserstein ambiguity sets’ concept for its reusable contribution to distributionally robust optimization, and the ‘joint Wasserstein ambiguity sets’ open question as a critical limitation in handling temporal dependencies within model predictive control. No datasets were approved as none were sufficiently described as unique, reusable benchmarks.
Approved Concepts
- Time-Varying Wasserstein Ambiguity Sets: Provides a tractable method for handling distributional uncertainty in MPC by dynamically adjusting the ambiguity set based on real-time data dispersion.
Approved Open Questions
- Joint Wasserstein Ambiguity Sets: In VPPs with battery storage, decisions at time ‘k’ influence the state at ‘k+1’. If the controller does not account for the joint evolution of price distributions, it may implement overly conservative strategies that ignore profitable, multi-period arbitrage opportunities.
Links
Metadata & Links
- url
- https://arxiv.org/abs/2605.14642
- paper_id
- 2605.14642
- paper_source
- openalex
- domain
- time-series
- tags
- forecasting-horizonseasonality-handling
- architectures
-
- datasets
-
- concept_slugs
- time-varying-wasserstein-ambiguity-sets
- dataset_slugs
-
- skill
- TimeSeriesSkill
- created_at
- 2026-05-17T06:07:59Z