Scholarlink

<p>Effective pressure management is essential for improving service reliability, reducing leakage, and preventing structural failures in water distribution networks (WDNs). However, most existing optimization strategies rely on single-scenario assumptions and therefore perform poorly when confronted with real-world uncertainty, including variable demand, intermittent supply conditions, and compound crisis events. This study introduces an integrated regret-based optimization framework that combines pressure-driven hydraulic simulation, scenario-specific pressure reducing valve (PRV) optimization, and cross-scenario robustness evaluation. Fifteen operational and crisis scenarios-ranging from mild pressure deficits to severe intermittent and hybrid failures-were constructed to represent a broad and realistic envelope of operating conditions. For each scenario, a lightweight population-based local search algorithm identified optimal PRV locations and settings, which were subsequently evaluated across all scenarios using a full regret matrix. Results show that while scenario-specific solutions exhibit strong localized performance, many degrade sharply under mismatched conditions. In contrast, the regret-minimizing configuration (S10) achieves the lowest worst-case performance gap and stable pressure behavior across all scenarios. Sensitivity testing further confirms the robustness of S10 to parameter perturbations. The findings highlight the value of regret analysis as a practical, risk-averse decision-support tool for utilities, providing a computationally efficient and scalable foundation for resilient pressure management under deep operational uncertainty.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A rapid-response regret-based framework for robust pressure management in water distribution networks under operational uncertainty

  • 期刊论文

  • 发表日期：

    2026年4月27日

查看全文
  • Mohammadreza Alizadeh Tataki Afshar,
  • Mahdi Miri,
  • Narges Moghaddassi

摘要

Effective pressure management is essential for improving service reliability, reducing leakage, and preventing structural failures in water distribution networks (WDNs). However, most existing optimization strategies rely on single-scenario assumptions and therefore perform poorly when confronted with real-world uncertainty, including variable demand, intermittent supply conditions, and compound crisis events. This study introduces an integrated regret-based optimization framework that combines pressure-driven hydraulic simulation, scenario-specific pressure reducing valve (PRV) optimization, and cross-scenario robustness evaluation. Fifteen operational and crisis scenarios-ranging from mild pressure deficits to severe intermittent and hybrid failures-were constructed to represent a broad and realistic envelope of operating conditions. For each scenario, a lightweight population-based local search algorithm identified optimal PRV locations and settings, which were subsequently evaluated across all scenarios using a full regret matrix. Results show that while scenario-specific solutions exhibit strong localized performance, many degrade sharply under mismatched conditions. In contrast, the regret-minimizing configuration (S10) achieves the lowest worst-case performance gap and stable pressure behavior across all scenarios. Sensitivity testing further confirms the robustness of S10 to parameter perturbations. The findings highlight the value of regret analysis as a practical, risk-averse decision-support tool for utilities, providing a computationally efficient and scalable foundation for resilient pressure management under deep operational uncertainty.