Research Of Several Issues In Pipeline Guided Wave Inspection

Guided wave inspection technology has great value in pipeline integrity monitoring forits high speed, long detection distance, comprehensive detection range as well as the ability todetect potential leak defects.Study of the signal processing techniques will improve the performance of defectsidentification and location. In this thesis, pipeline guided wave inspection theory anddeconvolution theory are reviewed，and two aspects of research are carried out in order tostudy guided wave signal processing technique based on deconvolution.Signal processing technology based on deconvolution can be used to improve theresolution of the detection signal. In this thesis, structure information of pipeline together withthe detection signal constitute the constrained deconvolution. Numerical simulation andexperimental analysis show that: structure constrained deconvolution has better anti-noiseability, good stability, higher resolution, and better defect locating accuracy, furthermore, ithas no requirement on the phase of wavelet.Signal processing algorithm will be affected by frequency dispersion of the signal,especially in long range pipeline guided wave inspection. In this thesis, dispersion simulationalgorithm is applied to research the influences of dispersion on statistical deconvolution anddeterministic deconvolution. Simulation results show that, both statistical deconvolution anddeterministic deconvolution are sensitive to dispersion of the signal. In order to applydeconvolution algorithm to the signal processing of long range pipeline guided waveinspection, dispersion compensation algorithm is applied to preprocess the signal. Numericalsimulation shows that: dispersion of the signal is eliminated to a certain extent by dispersioncompensation; and the stability of deconvolution algorithm is improved. Dispersioncompensation provides the possibility of applying deconvolution method to long range pipeline guided wave inspection.