Research On Sparse Data Acquisition Method Of Coded Ultrasonic Based On FPGA And Its Application For Plate Weld Defect Imaging

The plate structure is widely used in ship manufacturing,and the plates are connected by welding technology.The quality of the weld seam not only affects the quality of the ship,but also directly affects the safety of the ship.The welding seam will inevitably produce defects during the welding process.Accurate detection of weld defects and imaging is the important way to guarantee the quality of the welding seam.Ultrasonic phased array inspection technology has been widely used in the detection of such welds,but due to the limitations of this technology in the ultrasonic excitation method and data reception method,it is impossible to receive more abundant information about the defect.The imaging method based on FMC-TFM can obtain rich echo information,but restricts the real-time nature of data processing because of the large data.At the same time,these detection methods generally use pulsed ultrasonic excitation,and under the same excitation voltage the echo signal noise is relatively small,which is not conducive to the quantitative evaluation of defects.Therefore,the thesis uses coded ultrasound as the signal carrier and applies the sparse sampling method based on the finite rate of innovation to the data collection link,which greatly reduces the amount of collected data.Considering the time efficiency of sparse data during reconstruction,it is implemented on FPGA parameter reconstruction real-timely to image the defects by the sparse data reconstruction parameters,which provides a new method for the detection of weld defects.The thesis summarizes the current status and characteristics of non-destructive testing methods and technologies for sheet metal weld defects,and determines the research ideas and overall framework of this thesis in view of the limitations of current ultrasonic testing and imaging methods.Introduce the principle of coded ultrasonic defect detection,and the method and performance characteristics of coded ultrasonic generation are studied.In this thesis,binary frequency modulation coding is used as the excitation signal for the detection of plate weld defects.After the detection echo is matched filtered and Hilbert transformed,it is sparsely sampled based on the finite rate of innovation to obtain sparse data.Using FPGA to realize the parameter reconstruction algorithm of echo signal,which improves the time efficiency of parameter reconstruction.Then the defect acoustic characteristic parameters calculated by reconstruction are applied to defect imaging to obtain the FMC-TFM image of defect according to sparse data.The coding ultrasonic testing system for the defect of the plate welding seam was built to verify the performance of the proposed method.The main research work and conclusions are as follows:(1)Analyzed the technical status,basic principles and methods of pulsed ultrasonic inspection of plate welds.Based on binary frequency modulation coding as the excitation signal,analyze the principle of sparse sampling of coded ultrasonic,and formulated an overall plan which is using binary coded ultrasonic to detect the defect of plate weld.(2)Analyze the principle of sparse sampling of encoded ultrasound,and use the zeroing filter algorithm to reconstruct the characteristic parameters of the obtained sparse data to obtain the delay and amplitude parameters of the echo signal.A rapid reconstruction method of sparse data feature parameters based on FPGA is designed,and a sparse data parameter extraction experiment is carried out.Using a complete defect echo signal as an example,the time efficiency of parameter reconstruction is improved by about 19 times compared with Matlab reconstruction on a computer.(3)According to Fermat's theorem and Snell's law,a mathematical method to calculate the sound path of ultrasonic waves propagating inside the wedge and the detected object is given,so as to obtain the time index value required by the all-focus imaging algorithm.Combining the feature parameters extracted from sparse data,an all-focus imaging algorithm based on sparse data is proposed,which uses the delay and amplitude parameters in sparse data to perform all-matrix all-focus imaging.The simulation analysis of the sparse data all-focus imaging method was carried out.Compared with the conventional data all-focus imaging method by simulation.When the amount of data was significantly reduced,the defect image with the same quality as the conventional imaging method was obtained.(4)According to the characteristics of weld defects,a coded ultrasonic testing and testing system was built,and use 16-array linear array ultrasonic probe detect five types of weld defects which include transverse cracks,lack of fusion,lack of penetration,air holes and slag inclusions.The echo wave signal was sparsely sampled and reconstruct the parameter and five defects in the weld were imaged and compared with the defect detection images of the phased array.The sparse data was used to image the five defects of lateral cracks,non-fusion,non-welding,air holes and slag inclusions in the test.The imaging quality evaluation parameters API were 4.32,5.36,4.78,5.42 and 6.23 respectively.Using coded ultrasonic sparse sampling data to detect weld defects provides a reference.