压力容器2025，Vol.42Issue(9)：60-70,11.DOI:10.3969/j.issn.1001-4837.2025.09.008
基于导波频散回溯与稀疏贝叶斯学习的金属筒体结构裂纹定位方法研究
Study on crack localization method for metal cylindrical structures based on guided wave dispersion backtracking and sparse Bayesian learning
摘要
Abstract
To address the signal distortion caused by dispersion effects in ultrasonic guided wave testing and the tendency of traditional imaging algorithms to generate artifacts,a crack localization method integrating dispersion backtracking and sparse Bayesian learning is proposed.This method achieves dispersion compensation and wave packet travel distance extraction of guided wave signals based on dispersion backtracking,and introduces a dispersion curve correction strategy based on measured signals to rectify the deviation between the model and the actual structure.Meanwhile,sparse Bayesian learning is employed to fuse multipath information and solve for damage coordinates.Experimental validation on laboratory cylindrical structures and metal liners of hydrogen storage cylinders demonstrates that this method can achieve high-precision localization of cracks in metal cylindrical structures,with localization errors as low as 1.5%.Furthermore,combined with a square sensor array layout,it enhances adaptability to variations in defect orientation and expands its engineering applicability.This study provides a reliable technical approach for structural safety assessment of critical equipment such as pressure vessels.关键词
金属筒体结构/裂纹检测/超声导波/频散回溯/稀疏贝叶斯学习Key words
metal cylindrical structures/crack detection/ultrasonic guided waves/dispersion backtracking/sparse Bayesian learning分类
机械制造引用本文复制引用
王佳硕,周邵萍,罗瑞昌,程经纬,罗志..基于导波频散回溯与稀疏贝叶斯学习的金属筒体结构裂纹定位方法研究[J].压力容器,2025,42(9):60-70,11.基金项目
国家重点研发计划项目(2021YFB400800) （2021YFB400800）
国家自然科学基金项目(52175137,52205151) （52175137,52205151）
