Scholarlink

<p>This research aims to enhance the efficiency of integrated optimization design for solid launch vehicles (SLVs) concerning both internal and external ballistics. The solid rocket motor (SRM) model and trajectory simulation are established and optimized using both series and in parallel approaches. A comprehensive set of conventional serial design processes and integrated parallel design processes is proposed and compared. To improve the design efficiency of SLVs, a novel sequential approximate optimization (SAO) method is introduced to minimise the number of calls to the time-consuming original simulation models. This SAO method primarily relies on a combination of augmented radial basis functions (ARBF) and an inaccurate search strategy. ARBF effectively utilizes sample information from various local and global perspectives, with stronger generalization ability. The inaccurate search strategy incorporates an external elite pool to store optimal individuals and improves the global search ability by samples with certain randomness. The performance of the proposed methods has been verified by several popular optimization cases. Finally, compared to the conventional serial design process, the result shows that the proposed integrated parallel design process effectively reduced the initial mass of the objective three-stage SLVs and remarkably reduced the computational cost.</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

Integrated design of solid launch vehicles based on a novel sequence approximate optimization method

  • 期刊论文

  • 发表日期：

    2025年9月4日

查看全文
  • Shuaichao Ma,
  • Jiaxin Li,
  • Guosheng Li,
  • Yi Zhao,
  • Zeyang Xie,
  • Zeping Wu,
  • Jingwei Gao

摘要

This research aims to enhance the efficiency of integrated optimization design for solid launch vehicles (SLVs) concerning both internal and external ballistics. The solid rocket motor (SRM) model and trajectory simulation are established and optimized using both series and in parallel approaches. A comprehensive set of conventional serial design processes and integrated parallel design processes is proposed and compared. To improve the design efficiency of SLVs, a novel sequential approximate optimization (SAO) method is introduced to minimise the number of calls to the time-consuming original simulation models. This SAO method primarily relies on a combination of augmented radial basis functions (ARBF) and an inaccurate search strategy. ARBF effectively utilizes sample information from various local and global perspectives, with stronger generalization ability. The inaccurate search strategy incorporates an external elite pool to store optimal individuals and improves the global search ability by samples with certain randomness. The performance of the proposed methods has been verified by several popular optimization cases. Finally, compared to the conventional serial design process, the result shows that the proposed integrated parallel design process effectively reduced the initial mass of the objective three-stage SLVs and remarkably reduced the computational cost.