文献类型：期刊文献

中文题名：基于模态稀疏性与自适应正则化子空间追踪的结构应变响应重构

英文题名：Structural strain response reconstruction based on modal sparsity and adaptive regularized subspace pursuit

作者：Liang, Yi Wang, Bin Wang, Pengpeng Li, Yonghui Yin, Xiaoai Wu, Anjie

第一作者：梁意

机构：[1] School of Civil Engineering, Guizhou Institute of Technology, Guiyang, 550003, China; [2] Hunan Chuan Cheng Industrial Co., Ltd., Yiyang, 413500, China; [3] School of Transportation Engineering, Guizhou Institute of Technology, Guiyang, 550003, China

第一机构：贵州理工学院土木工程学院

年份：2026

卷号：45

期号：17

起止页码：55-66

外文期刊名：Zhendong yu Chongji/Journal of Vibration and Shock

收录：EI(收录号：20263221264502)

语种：中文

外文关键词：Iterative methods - Reconstruction (structural) - Safety engineering - Sensor networks - Structural dynamics - Vibrations (mechanical)

摘要：To address the high cost, difficult deployment, and sensor-failure risk of dense sensor networks, a sparse reconstruction method is proposed for full-field dynamic strain responses. Structural dynamics is integrated with compressive sensing theory in the proposed method. First, a sparse representation model of the structural response is established based on the modal superposition principle. Then, the temporal continuity of the response is used. An adaptive regularized subspace pursuit algorithm is developed based on this continuity. Sensor placement is further optimized using the reciprocity criterion and related strategies to improve reconstruction accuracy. Finally, the effectiveness of the proposed method is validated through numerical examples and vehicle-bridge vibration experiments. The mean relative percentage error of ARSP is 3. 86% . This value is lower than those of the subspace pursuit algorithm and the iterative hard thresholding algorithm, which are 5.65% and 23. 15%, respectively. This study provides a useful reference for structural dynamic monitoring under conditions of limited sensor availability. ? 2026 Chinese Vibration Engineering Society. All rights reserved.