Scholarlink

<p>A semiautomatic design system for river channel cross-sections is proposed that applies the concept of generative design to nature-oriented river management. The system processes a single representative cross-section of the current reach to generate multiple trapezoidal cross-sections according to the prescribed design parameters and in situ requirements. The cross-sections are then evaluated for flood safety, environmental diversity, and constructability by one-dimensional nonuniform hydraulic simulations, riffle-pool sequence segmentation, and soil volume cutting. As a case study, the design system was applied to two rural rivers with different hydraulic characteristics. The design system took approximately 20&#xa0;s to generate 100 cross-sections for each river, and the many options and indices helped with the unbiased selection of a reasonable cross-section. The most suitable cross-sections corresponded to those identified by previously developed empirical river restoration methods. For both rivers, the optimal cross-sections differed when the soil volume was considered. The proposed design system is expected to improve the efficiency and effectiveness of river channel design in the context of nature-oriented river 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

Semiautomatic design system for river channel cross-sections based on generative design

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年1月6日

查看全文
  • Takanori Kono,
  • Kazuaki Ohtsuki,
  • Kazufumi Hayashida,
  • Keigo Nakamura

摘要

A semiautomatic design system for river channel cross-sections is proposed that applies the concept of generative design to nature-oriented river management. The system processes a single representative cross-section of the current reach to generate multiple trapezoidal cross-sections according to the prescribed design parameters and in situ requirements. The cross-sections are then evaluated for flood safety, environmental diversity, and constructability by one-dimensional nonuniform hydraulic simulations, riffle-pool sequence segmentation, and soil volume cutting. As a case study, the design system was applied to two rural rivers with different hydraulic characteristics. The design system took approximately 20 s to generate 100 cross-sections for each river, and the many options and indices helped with the unbiased selection of a reasonable cross-section. The most suitable cross-sections corresponded to those identified by previously developed empirical river restoration methods. For both rivers, the optimal cross-sections differed when the soil volume was considered. The proposed design system is expected to improve the efficiency and effectiveness of river channel design in the context of nature-oriented river management.