CVaR-Guided Decision-Focused Learning and Risk-Triggered Re-Optimization for Two-Stage Robust Microgrid Operation

TL;DR

This paper introduces a CVaR-guided decision-focused learning framework with risk-triggered re-optimization for robust microgrid operation, improving tail-risk mitigation and computational efficiency.

Contribution

It develops a novel probabilistic forecasting and optimization approach that aligns load predictions with operational robustness using CVaR guidance and implicit differentiation.

Findings

Enhanced probabilistic forecasting accuracy on microgrid cases.

Reduced daily solution time by up to 91%.

Maintained near-optimal operational costs with less than 0.5% increase.

Abstract

Microgrid operation is highly vulnerable to short-term load uncertainty, while conventional predict-then-optimize pipelines cannot fully align probabilistic forecasting quality with downstream robust scheduling performance. This paper proposes a CVaR-guided decision-focused learning and risk-triggered re-optimization framework for two-stage robust microgrid operation. A probabilistic load forecasting model first generates multi-quantile outputs, which are converted into prediction intervals to parameterize the load uncertainty set of the downstream two-stage robust optimization (TSRO) model. To improve forecasting reliability under difficult and high-risk operating conditions, a CVaR-guided forecasting objective is introduced to emphasize tail-sensitive samples. To further close the forecast-decision gap, a convex regularized surrogate TSRO model and a smooth regret loss are developed, enabling downstream operational feedback to be propagated to the forecasting model through KKT-based implicit differentiation. For online deployment, a risk-triggered re-optimization mechanism selectively re-solves the remaining-horizon TSRO only when the schedule mismatch becomes significant, avoiding unnecessary online computation. Case studies on modified IEEE 33-bus and 69-bus microgrids demonstrate superior probabilistic forecasting accuracy, operational economy, and tail-risk mitigation over benchmark methods, while preserving near-full-re-optimization performance with less than 0.5% higher operating cost and up to 91% lower daily solution time.