Compressed Sensing Method for Health Monitoring of Pipelines Based on Guided Wave Inspection

Zhe Wang,Songling Huang,Shen Wang,Shuangyong Zhuang,Qing Wang,Wei Zhao
DOI: https://doi.org/10.1109/tim.2019.2951891
IF: 5.6
2019-01-01
IEEE Transactions on Instrumentation and Measurement
Abstract:The pipeline in-service needs to be inspected in a certain period to master its structural health status. An ultrasonic guided wave, which can propagate along pipelines with less energy loss, provides an efficient method for long-term in situ inspection. The guided waves can detect both corrosion and cracks existing in structures. To overcome the problem of huge amounts of data and to maintain defect identification accuracy, the compressed sensing method for guided wave inspection is proposed. The compression process is essentially a scheme of analog to information conversion to compress the signal. It is accomplished by random demodulation and the equivalent sampling rate below the Nyquist rate helps to save most of the storage. Compressed data are recovered to the sparse spatial domain based on the constructed dictionary from a guided wave propagation model. To verify the effectiveness of the proposed method, both numerical simulations and experimental investigations are conducted. The results indicate the availability of compression and high accuracy of defect location after recovery. The influences of different compression schemes and compression ratios are further analyzed. In addition, the comparisons with direct recovery without compression and traditional analysis methods demonstrate the advantageous performance of the proposed method.
What problem does this paper attempt to address?