Distributionally Robust Reserve Optimization Method for Multi-Region Interconnected Power Systems Incorporating Risk Management

To address cross-regional power imbalances and privacy vulnerabilities in conventional dispatch frameworks, this paper proposes a risk-aware reserve capacity optimization method for multi-region interconnected power systems. A data-driven uncertainty set is developed by integrating historical prediction error distributions, where risk deficit calibration dynamically balances conservatism and economic objectives under renewable energy volatility. The framework establishes a two-stage robust optimization model coordinating reserve allocation and real-time operational risks, effectively handling renewable intermittency and demand fluctuations. To decentralize decision-making, an alternating direction method of multipliers (ADMM)-based architecture is designed, enabling regions to autonomously optimize local resources through boundary power exchange coordination autonomously, thereby eliminating centralized data sharing and reducing inter-regional communication dependencies. A nested column-and-constraint generation (C&CG) algorithm is introduced to resolve mixed-integer nonlinear challenges, decomposing master subproblems iteratively to enhance computational tractability while managing discrete commitment variables. Theoretical analysis demonstrates the method's robustness in maintaining feasibility under worst-case uncertainty realizations.