一种基于压缩感知的导波场重构方法

导波场分析法可有效识别和表征结构损伤,但受限于奈奎斯特采样定律,全波场信号的采集耗时严重。为提高全波场信号的获取效率,现有方法依靠压缩感知和激光多普勒测振技术,以少量空间测点信号重构出原始波场,然而该类方法依赖具有一定随机性的抖动采样策略,在商用激光多普勒测振系统上不便实现。为解决这一问题,该文提出了一种基于均匀稀疏采样策略的导波场重构方法,并搭建压电片激励/扫描式激光多普勒测振仪传感(PZT激励/SLDV传感)实验平台,对含人工损伤的铝板进行实验验证。实验结果表明,该方法可将空间测点数减少到奈奎斯特采样点数的10%以下,并在导波场重构精度及损伤成像精度方面达到与抖动采样策略的同等水平,且提高了导波场分析法的实用性和工作效率。

A Guided Wave Field Reconstruction Method Based on Compressed Sensing

The guided wave field analysis method can effectively identify and characterize the structural damage, but due to the limitation of Nyquist sampling law, the acquisition of full wave field signals is often time consuming. In order to improve the acquisition efficiency of full wave field signals, the existing methods rely on compressed sensing technology and laser Doppler vibration measurement technology to reconstruct the original wave field with a small number of spatial signals. However, these methods rely on a random jittered sampling strategy, which is difficult to achieve in commercial laser Doppler vibration measurement system. In order to solve this problem, a guided wave field reconstruction method based on uniform sparse sampling strategy is proposed in this paper, and a PZT excitation/SLDV sensing experimental platform is constructed to verify this method on an aluminum plate with artificial damage. The experimental results show that the proposed method can reduce the number of space measurement points to less than 10% of Nyquist sampling points, and achieve the same level of reconstruction accuracy and damage imaging accuracy as the existing jittered sampling strategy, which can greatly improve the practicability and efficiency of guided wave field analysis method.