Research On The Mechanism And Technology Of Rapid Detection Of Laser Ultrasound Internal Defects Based On Dichotomous Search And Compression Perception

In laser ultrasonic nondestructive testing,the time-of-flight scattering method and synthetic aperture focusing can detect internal defects in the sample to be tested,but because the location of the defects within the sample to be tested is unknown,a large area scan is required in the process of defect localisation,resulting in a large number of spatial sweep points and a serious time-consuming inspection.In order to solve the above serious time-consuming problems,this paper proposes a combination of laser ultrasonic NDT techniques with dichotomous search and compressive sensing methods.The method selects sparse sampling points by the dichotomous search method and uses the signal sparsity property brought by the compressed sensing method to extract the arrival time of reflected signals from in vivo transverse perforation defects in order to achieve the purpose of accelerated in vivo defect localisation.Firstly,the principles of the correlation coefficient method and the feature extraction method are introduced and investigated for the target signal extraction problem.And then,based on the analysis of the results of the measured ultrasonic signals of defects at different locations under the two methods,the surface and in vivo transverse perforation defect feature extraction problem is solved.For the ultrasonic signal compression problem,the principle of dichotomous search algorithm and the principle of compression perception algorithm are investigated.Based on these studies,an accelerated technique for defect localisation based on dichotomous search and compressive sensing is proposed.As the base used in the compressive sensing method in this paper is wavelet base,the selection of wavelet base is studied and analysed in order to accurately output the arrival time of the transverse wave reflection signal of the in vivo transverse hole defect,and the problem of feature signal recognition error is solved.Then,numerical simulations are used to study the localisation and acceleration algorithm for the localisation of in vivo transverse hole defects in the one-dimensional case and to solve the localisation and acceleration problem of in vivo transverse hole defects.The excitation detection method is also investigated to select the suitable excitation detection method for this paper.For the defect localisation acceleration problem,the dichotomous search method and the compressed sensing method are used to solve the transverse localisation acceleration problem of the in vivo transverse hole defect,and then the depth information of the in vivo transverse hole defect is solved by the time flight scattering method based on the above acceleration techniques.The localisation results show that the lateral position is offset from the actual position by 6% and the transverse position is offset from the actual position by 8%.Finally,an experimental approach to the problem of locating defects in internal transverse holes at different locations was used to achieve accelerated localisation of defects in internal transverse holes.A laser ultrasound scanning device for the detection of internal defects was constructed,using a pulsed laser for laser excitation and a Doppler vibrometer for non-contact detection of the ultrasound,and a dichotomous search and compression-aware rapid detection of internal defects by moving the sample at a fixed excitation detection distance.To address the problem of feature signal extraction,the arrival time of the reflected signal of the transverse wave of the in vivo transverse hole defect in the measured ultrasound signal is output using the same processing method as in the simulation,and the problem of locating the transverse and depth position of the in vivo transverse hole defect is solved.The experimental results show that the proposed localisation acceleration technique shortens the time required to locate transverse in vivo defects compared to point-by-point sweeping under the same experimental premise.The localisation results show that the lateral position deviates from the actual position by 1.6% and the lateral position deviates from the actual position by 3%.This paper proposes a compressed sensing algorithm based on the traditional dichotomous search and compressed sensing algorithms applicable to laser ultrasound line source excitation.In the experiments,the line source is used to circumvent the possible damage to the sample caused by point source excitation at the same energy.The algorithm is efficient in defect detection,requires less experimental equipment and environment,and can provide a new idea for accelerating the localisation of transverse perforation defects in vivo.