Scholarlink

<p>This study presents an effective application of the dragonfly algorithm (DA) to find the optimal placement of flow control valves (FCVs) in a water distribution network (WDN) for leakage reduction. Unlike previous work that used traditional meta-heuristic algorithms such as genetic algorithms (GA) or ant colony optimization (ACO), this research proposes and uses new meta-heuristic algorithm, named DA, for the WDN leakage control problem to improve search diversity and solution stability. The approach is applied to the Jowitt network as a case study, leading to a practical engineering solution with direct applicability for cost-effective leakage reduction. Here, the results were obtained under considering two operational conditions for the FCV (0.3 and 0.4) and outcomes are compared with GA. The results show that DA consistently outperforms GA, leading to minimum leakage values with faster convergence rates and lower standard deviation. In other words, for a setting value of 0.3, the DA leads to a minimum leakage value of 0.29, whereas the corresponding value obtained by the GA is 0.33. In addition, for a setting value of 0.4, the DA leads to a minimum leakage value of 0.33, in contrast to the corresponding value of 0.34 obtained by the GA. This demonstrates that the DA is slightly more efficient to minimize leakage values. This fact demonstrates the algorithm’s potential as a reliable, efficient, and implementable decision-support tool for sustainable WDN management.</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

Reducing leakage using the dragonfly algorithm to find the optimal placement of flow control valves in water distribution networks

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月14日

查看全文
  • Mehrtash Eskandaripour,
  • Ramtin Moeini,
  • Mohamad Reza Najarzadegan

摘要

This study presents an effective application of the dragonfly algorithm (DA) to find the optimal placement of flow control valves (FCVs) in a water distribution network (WDN) for leakage reduction. Unlike previous work that used traditional meta-heuristic algorithms such as genetic algorithms (GA) or ant colony optimization (ACO), this research proposes and uses new meta-heuristic algorithm, named DA, for the WDN leakage control problem to improve search diversity and solution stability. The approach is applied to the Jowitt network as a case study, leading to a practical engineering solution with direct applicability for cost-effective leakage reduction. Here, the results were obtained under considering two operational conditions for the FCV (0.3 and 0.4) and outcomes are compared with GA. The results show that DA consistently outperforms GA, leading to minimum leakage values with faster convergence rates and lower standard deviation. In other words, for a setting value of 0.3, the DA leads to a minimum leakage value of 0.29, whereas the corresponding value obtained by the GA is 0.33. In addition, for a setting value of 0.4, the DA leads to a minimum leakage value of 0.33, in contrast to the corresponding value of 0.34 obtained by the GA. This demonstrates that the DA is slightly more efficient to minimize leakage values. This fact demonstrates the algorithm’s potential as a reliable, efficient, and implementable decision-support tool for sustainable WDN management.