脆弱な都市の災害対策支出を最適化する新インフラモデル(New Infrastructure Model Prioritizes Disaster Spending for Vulnerable Cities)

2026-06-23 ヒューストン大学(UH)

米国ヒューストン大学の研究チームは、大規模災害発生後に限られた資源をどのように優先配分すべきかを支援する新たな意思決定モデルを開発した。災害時には、電力、通信、交通、医療、水道など相互に依存する重要インフラが同時に被害を受けることが多く、復旧の優先順位を誤ると社会全体の回復が遅れる。研究では、インフラ間の依存関係や社会的影響を考慮した数理モデルを構築し、どの施設やサービスを優先的に復旧すれば地域全体の機能回復を最大化できるかを評価した。モデルは単なる物理的損傷だけでなく、住民への影響や復旧資源の制約も考慮するため、より実践的な災害対応計画の策定に活用できる。研究者らは、この手法がハリケーン、洪水、地震など多様な災害への適用を可能にし、地域社会のレジリエンス向上に貢献するとしている。今後は自治体やインフラ事業者による防災計画や復旧戦略の高度化への活用が期待される。

脆弱な都市の災害対策支出を最適化する新インフラモデル(New Infrastructure Model Prioritizes Disaster Spending for Vulnerable Cities)
A new mathematical model developed at UH shows that targeted investments in a small number of critical assets can significantly improve infrastructure resilience and disaster recovery.

<関連情報>

インフラのレジリエンスに関する確率制約付き最適化:公益事業主導の予算配分フレームワーク Chance-constrained optimization of infrastructure resilience: A utility-driven budget allocation framework

Tugce Uslu Aktas, Gino J. Lim, Jian Shi
Computers & Industrial Engineering  Available online: 23 May 2026
DOI:https://doi.org/10.1016/j.cie.2026.112120

Highlights

  • A chance-constrained model is developed to enhance system resilience.
  • The framework is demonstrated to be applicable under various distributional assumptions.
  • A successive convex approximation is applied to non-convex utility functions.
  • Implementation is illustrated via real-world power and transport cases.
  • An efficient budget allocation plan to maximize resilience is obtained.

Abstract

This paper presents a novel chance-constrained model to enhance infrastructure resilience in a stochastic environment under budget constraints. Our goal is to maximize resilience by accounting for the system’s robustness and recovery capability, both of which are essential indicators of a resilient system. The uncertainty in functional degradation and recovery time of components is considered due to the uncertain intensity of adverse occurrences. The framework accommodates multiple distributional assumptions, including normal and uniform distributions, and integrates both linear and nonlinear utility functions to determine optimal trade-offs between reducing functional degradation and improving recovery. In addition, to efficiently solve the nonconvex utility formulation, a successive convex approximation algorithm is implemented, which significantly improves computational performance. The model is validated through three case studies (an illustrative system, a power transmission network, and a transportation network), demonstrating its applicability across different infrastructures. The results highlight the framework’s ability to generate efficient budget allocation plans that maximize resilience while accounting for uncertainty in extreme events.

2100総合技術監理一般
ad
ad
Follow
ad
タイトルとURLをコピーしました