Proactive Robust Hardening of Resilient Power Distribution Network: Decision-Dependent Uncertainty Modeling and Fast Solution Strategy

Donglai Ma, Xiaoyu Cao, Bo Zeng, Qing-Shan Jia, Chen Chen, Qiaozhu Zhai, Xiaohong GuanIEEE Trans Autom. Sci. Eng 2025 · DOI · OpenReview

TL;DR

This paper introduces a novel "decision-dependent uncertainty" (DDU) model for optimizing power distribution network hardening against extreme weather events. By dynamically linking hardening decisions to potential failure scenarios and incorporating load variability, their DDU-RSO framework offers a more realistic resilience assessment. They also developed an enhanced parametric column-and-constraint generation (P-C&CG) algorithm that significantly speeds up the solution process, achieving orders-of-magnitude improvements over existing methods.

What Problem Does This Paper Address?

Power grids are increasingly vulnerable to extreme weather events, leading to widespread outages. Traditional approaches to "harden" these grids (e.g., reinforcing lines, burying cables) often use robust optimization (RO) that assumes a fixed set of potential failures. However, hardening certain components can actually make other failure scenarios impossible or less likely. This disconnect between hardening decisions and the assumed contingencies leads to an incomplete or overly conservative resilience strategy. This paper addresses this by proposing a "decision-dependent uncertainty" (DDU) model, which explicitly accounts for how hardening choices influence the likelihood and impact of different failure scenarios. Furthermore, it integrates the randomness of power loads and energy storage availability, which are crucial for post-event recovery.

Methods

The core of the paper is the Decision-Dependent Uncertainty-based Robust-Stochastic Optimization (DDU-RSO) model. This is a trilevel optimization framework:

  1. Outer Level (Defender): Determines proactive hardening decisions (which lines/generators to reinforce).
  2. Middle Level (Attacker): Identifies the worst-case contingency scenario, considering how the hardening decisions affect these scenarios (the DDU set).
  3. Inner Level (Restoration): Optimizes operational corrective measures (like scheduling distributed generators and energy storage) to minimize load shedding under the worst-case scenario, considering load and storage uncertainties via stochastic programming.

To solve this complex DDU-RSO model, the authors developed an enhanced Parametric Column-and-Constraint Generation (P-C&CG) algorithm. This algorithm leverages network structural information and "resilience importance indices" to improve convergence and significantly reduce computation time. The resilience importance index quantifies the impact of a single line/DG failure on system resilience, helping to focus the optimization on critical components.

Datasets

The paper evaluates its proposed method on two standard test distribution networks:

  • IEEE 33-bus test distribution network
  • 118-bus test distribution network

These networks are augmented with distributed generators (DGs) and energy storage systems (ESSs) to simulate realistic scenarios.

Key Insights and Ideas

  • Decision-Dependent Uncertainty (DDU) is Crucial: The paper strongly argues that ignoring the feedback loop between hardening decisions and contingency scenarios leads to suboptimal resilience planning. DDU captures a more realistic system behavior.
  • Holistic Uncertainty Modeling: Integrating both endogenous (contingency) and exogenous (load, storage) uncertainties provides a more comprehensive picture of system vulnerability and recovery potential.
  • Resilience Importance Indices Enhance Efficiency: By quantifying the importance of network components for resilience, these indices help the P-C&CG algorithm converge faster and avoid the "multi-solution dilemma" often encountered in complex optimization problems.
  • DDU-RSO can be More Efficient than DIU-RSO: Counter-intuitively, modeling DDU and solving it directly with their enhanced P-C&CG can be faster than reformulating it into a simpler, but less accurate, decision-independent uncertainty (DIU) model.

Key Results

  • Improved Resilience: The DDU-RSO framework, aided by hardening strategies, significantly reduces load shedding during extreme events. For the 33-bus system, load shedding was reduced from over 54% in the baseline case to 25.8% with hardening.
  • Computational Speedup: The enhanced P-C&CG algorithm achieved orders-of-magnitude faster solution times compared to existing methods like basic C&CG and modified Benders decomposition (BD). For instance, on the 118-bus system, the enhanced P-C&CG reduced solution time by up to 86.7% compared to basic C&CG and significantly outperformed BD, which often failed to converge within the time limit.
  • Scalability: The algorithm demonstrated good scalability, performing effectively on both the 33-bus and the larger 118-bus test systems.

Limitations and Future Work

  • Mixed-Integer Recourse: The paper notes that while their current framework handles stochastic programming in the innermost level, extending it to include mixed-integer recourse variables (e.g., for network reconfiguration or mobile resource scheduling) would convert the innermost problem into a Mixed-Integer Program (MIP). This would require further algorithmic development, potentially using nested C&CG, to maintain computational efficiency.
  • Generalizability of DDU: While the DDU model is presented for hardening, its application to other resilience enhancement measures (like demand response, network reconfiguration, mobile resources) is suggested as future work.
  • Computational Complexity: Although the enhanced P-C&CG is highly efficient, the paper acknowledges that master problems can become intractable with an exponential increase in complexity. However, they argue their parametric cuts are more robust than Benders-type dual cuts.

Summary updated

Abstract

To address the power system hardening problem, traditional approaches often adopt robust optimization (RO) that considers a fixed set of concerned contingencies, regardless of the fact that hardening some components actually renders relevant contingencies impractical. In this paper, we directly adopt a dynamic uncertainty set that explicitly incorporates the impact of hardening decisions on the worst-case contingencies, which leads to a decision-dependent uncertainty (DDU) set. Then, a DDU-based robust-stochastic optimization (DDU-RSO) model is proposed to support the hardening decisions on distribution lines and distributed generators (DGs). Also, the randomness of load variations and available storage levels is considered through stochastic programming (SP) in the innermost level problem. Various corrective measures (e.g., the joint scheduling of DGs and energy storage) are included, coupling with a finite support of stochastic scenarios, for resilience enhancement. To relieve the computation burden of this new hardening formulation, an enhanced customization of parametric column-and-constraint generation (P-C&CG) algorithm is developed. By leveraging the network structural information, the enhancement strategies based on resilience importance indices are designed to improve the convergence performance. Numerical results on 33-bus and 118-bus test distribution networks have demonstrated the effectiveness of DDU-RSO aided hardening scheme. Furthermore, in comparison to existing solution methods, the enhanced P-C&CG has achieved a superior performance by reducing the solution time by a few orders of magnitudes.
BibTeX
@article{ma2025proactive,
  author = {Donglai Ma and Xiaoyu Cao and Bo Zeng and Qing-Shan Jia and Chen Chen and Qiaozhu Zhai and Xiaohong Guan},
  title = {{Proactive Robust Hardening of Resilient Power Distribution Network: Decision-Dependent Uncertainty Modeling and Fast Solution Strategy}},
  journal = {IEEE Trans Autom. Sci. Eng.},
  year = {2025},
  doi = {10.1109/tase.2025.3564483},
}
Figure