超声波传感器
叠加原理
声学
无损压缩
滤波器(信号处理)
信号(编程语言)
平滑的
能量(信号处理)
傅里叶变换
计算机科学
噪音(视频)
算法
数学
人工智能
物理
计算机视觉
数学分析
统计
数据压缩
图像(数学)
程序设计语言
作者
Geolei Zheng,Weigao Zhao,Yimei Tian,Chuntong Liu,Bo Zheng
标识
DOI:10.1088/1361-665x/ab8b2f
摘要
Ultrasonic guided waves have been widely used for structural-damage localization in recent decades, whose effectiveness depends on whether the damage on a structure reflects a wave pulse. The basic principle can be described as follows: the ratio between the times of flight (TOFs) of the excitation and damage signals is equal to that between their distances of flight. Because a number of disturbing factors occurred during the data-acquisition process, a large number of noise signals present in the experiment could cause a low signal-to-noise ratio (SNR). In addition, whereas the traditional filtering methods improved the SNR, the effective signals were weakened. To deal with this situation, an ultrasonic lossless filtering method based on the multi-coverage technique applied to a blasting seismic wave and the superposition technique to ultrasonic guided wave was first proposed to enhance the SNR. Compared with the traditional filtering techniques, the proposed method achieves better performance in that the effective signal remains unaltered and the noise level is reduced. The Wigner–Ville distribution (WVD) method, which could reflect the true energy distribution within the time and frequency domains, was applied to calculate the TOF of the signals. Finally, the damage localization of the structure was realized according to the abovementioned principle. Compared with the location result without a filter, the maximum WVD value becomes more perceivable, and the positioning accuracy is clearly improved. Compared with the location results using the Fourier and polynomial-smoothing filtering methods, the lossless filtering method allows an effective signal to remain unaltered. Finally, we can conclude that the proposed ultrasonic lossless filtering method can improve the SNR under the condition in which the effective signal remains unaltered, and the SNRs become higher and higher with the increase of superimposition times. The combination of ultrasonic lossless filtering method and the WVD method can perfectly realize the damage location with a high accuracy. Several damage-localization experiments demonstrate that the ultrasonic lossless filtering method is reliable.
科研通智能强力驱动
Strongly Powered by AbleSci AI