航空科学技术2025，Vol.36Issue(5)：51-58,8.DOI:10.19452/j.issn1007-5453.2025.05.006
基于频谱压缩感知和核极限学习机的柔性复合材料结构冲击定位
Impact localization of Flexible Composite Material Structure Based on Spectrum Compression Sensing and Kernel Extreme Learning Machine
摘要
Abstract
Flexible composite materials structures,owing to their advantages such as light weight,good repeatability of deployment and retraction,and simple principles of deployment and retraction,are widely applied in aerospace vehicle structures.These structures may suffer from impact loads and thus generate damages during their service.In order to address the issue of impact localization on flexible composite structures,an impact monitoring system based on fiber bragg gratings was constructed.The time and frequency domain characteristics of the impact signals is analyzed,taking advantage of the sparsity of spectral signals and compressive sensing to reconstruct signals under conditions far below the Nyquist sampling rate,an impact localization method for flexible composite material structures based on frequeny spectral compressed sensing and kernel extreme learning machine(KELM)is proposed.This study focuses on the sparse representation and compressed measurement of the frequency spectrum for the impact response signals of flexible composite materials.The observation vector,obtained by compressing and reducing the dimensionality of the impact response signal spectrum,is used as a feature.A subtraction-average-based optimizer(SABO)is employed to optimize the kernel extreme learning machine(KELM)for identifying the location of impact loads on flexible composite materials.Fifteen test sample points are randomly selected within a 200mm x 400mm monitoring area on the flexible composite material structure for impact localization.Experimental results show that the average localization error is 18.83mm,providing a reliable method for impact localization on flexible composite structures.关键词
冲击定位/柔性复合材料结构/光纤光栅/压缩感知/核极限学习机Key words
impact localization/flexible composite structure/fiber bragg gratings/compression sensing/kernel extreme learning machine分类
计算机与自动化引用本文复制引用
喻俊松,刘君,彭子鹏,干灵辉,万生鹏..基于频谱压缩感知和核极限学习机的柔性复合材料结构冲击定位[J].航空科学技术,2025,36(5):51-58,8.基金项目
国家自然科学基金(62105139) （62105139）
航空科学基金(2022Z057056003) National Natural Science Foundation of China(62105139) （2022Z057056003）
Aeronautical Science Foundation of China(2022Z057056003) （2022Z057056003）
