系统工程与电子技术 ›› 2025, Vol. 47 ›› Issue (7): 2283-2303.doi: 10.12305/j.issn.1001-506X.2025.07.22

• 系统工程 • 上一篇    

基于结构冗余传感器配置与灰狼优化算法的无人机可诊断性优化设计

谷旭平, 史贤俊   

  1. 海军航空大学岸防兵学院, 山东 烟台 264001
  • 收稿日期:2024-07-02 出版日期:2025-07-16 发布日期:2025-07-22
  • 通讯作者: 谷旭平
  • 作者简介:谷旭平 (1996—), 男, 博士研究生, 主要研究方向为可诊断性、可重构性、测试性
    史贤俊 (1968—), 男, 教授, 博士, 主要研究方向为可诊断性、可重构性、测试性

Diagnosability optimal design of UAV based on structurally redundant sensor configurations with grey wolf optimization algorithm

Xuping GU, Xianjun SHI   

  1. College of Coastal Defense Force, Naval Aviation University, Yantai 264001, China
  • Received:2024-07-02 Online:2025-07-16 Published:2025-07-22
  • Contact: Xuping GU

摘要:

为提高无人机的可诊断性, 提出基于结构冗余传感器配置与灰狼优化算法的可诊断性优化设计策略。首先, 为弥补结构分析在衡量故障诊断难易程度的缺陷, 提出基于Wasserstein距离的可诊断性评价方法。其次, 设计结构冗余传感器配置算法, 以最低的传感器配置代价实现系统可诊断性最大化。最后, 提出基于灰狼优化算法的可诊断性优化设计策略, 在满足系统可诊断性定性和定量需求的前提下, 最小化诊断系统设计代价。并基于固定翼无人机结构模型, 利用所提算法, 以最小的传感器优化配置代价, 使得系统可检测率和可隔离率达到100%。此外, 基于定性评价的优化策略, 使得诊断代价缩减83%, 较其他算法节省2%~15%;并基于定量评价的优化策略, 使得诊断代价缩减90%, 较其他算法节省0%~25%。

关键词: 结构分析, 可诊断性, 传感器配置, 灰狼优化算法

Abstract:

To improve the diagnosability of unmanned aerial vehicles (UAVs), this paper proposes an optimal design strategy for diagnosability based on structural redundant sensor configurations and the grey wolf optimization algorithm. Firstly, to make up for the defect of structural analysis in measuring the difficulty of fault diagnosis, a quantitative evaluation method of diagnosability based on Wasserstein distance is proposed. Secondly, a structural redundant sensor configuration algorithm is designed to maximize the system's diagnosability with the lowest sensor configuration cost. Finally, a diagnosability optimization design strategy based on the gray wolf optimization algorithm is proposed to minimize the design cost of the diagnosability system while meeting the qualitative and quantitative diagnosability requirements. Based on the fixed UAV structural model, using the proposed algorithm, the system detectability and isolation rate reaches 100% with the minimum cost of optimal sensor configuration. The optimization strategy based on qualitative evaluation makes the diagnostic cost shrink by 83%, which is a saving of 2% to 15% compared with the other algorithms. The optimization strategy based on quantitative evaluation makes the diagnostic cost shrink by 90%, which is a saving of 0% to 25% compared with the other algorithms.

Key words: structural analysis, diagnosability, sensor configuration, grey wolf optimization algorithm (GWO)

中图分类号: