Research On Composite Non-destructive Detection Technology Of ECT And EMAT

With the rapid development of modern industry and constantly improvement of important equipment safety monitoring and security assurance requirements, non-destructive detection technology will face more difficulties and challenges, and sometimes a single detection method has been unable to meet the actual testing requirements. Effectively combining the advantages of different detection methods to accomplish more complex detection tasks has important practical significance. In this paper, supported by the National Natural Science Foundation of China, reasearch has been carried out on the development of ECT/EMAT composite detection system, signal processing and defect detection technology, implementing the ECT/EMAT noncontact defect detection on plate-like conductor.The main work and innovations are as follows:(1) To meet the practical needs of nondestructive detection of defects with different depth locations in plate-like conductors. An ECT/EMAT composite detection method was proposed. The composite principles and methods of the ECT/EMAT systems at mechanism level and system level have been researched. The feasibility of system composite method has been analysised from the composite mechanism, detection method and signal processing aspects. The analysis shows that ECT and EMAT detection techniques have strong complementarities, and the system level composite detection method can effectively improve the scanning efficiency and detection capability for different defects at different depth locations.The principle and feasibility research provide the theoretical basis for the development of the dual-mode system.(2) Based on the research and analysis of composite detection mechanism and feasibility, an ECT/EMAT composite detection prototype system has been developed. The signal generation and detection signal processing technology of ECT device, high-frequency pulse excitation and signal reception technology of EMAT device, and the development technology of composite detection software system have been researched. In the development of ECT device, power amplifier circuit is inducted to increase the excitation power, and the probe balance circuit is added to reduce ambient noise. And in the development of EMAT prototype device, excitation circuit, signal receiving circuit and impedance matching circuit have been designed. In the development of software, an ECT/EMAT real-time detection software based on Labview and Matlab environment has been developed, realizing the signal real-time acquisition, signal processing, data storage and defect detection functions. (3) For the adverse effect on defect detection from the shake of lift-off (the distance between the probe and the measured object), signal noise and other factors, signal preprocessing technology of ECT and EMAT detection signal has been researched. Based on the analysis and research on eddy current testing probe impedance model, the probe impedance has been proved to be proportionally changed with the variation of lift-off. Combined with the proportional changing characteristics and radial base interpolation, a novel lift-off invariance transformation method has been proposed, the detection signals can be proportionally mapped to the zero lift-off signal, and the influence of lift-off fluctuations can be better suppressed. In additon, the wavelet analysis method has been utilized to suppress and reduce the EMAT signal noise. The effectiveness of the researched method has been validated by the experimental testing.(4) The plate-like conductor detection method based on ECT and EMAT signals has been researched. The Support Vector Regression (SVR) algorithm has been adopted to analysis the quantitation of defects,.And in order to improve the training performance of small sample inversion model, the particle swarm optimization (PSO) algorithm has been used to optimize the SVR parameters; In order to play respective advantages of ECT/EMAT detection method, a data fusion strategy based on threshold selection and machine learning algorithm has been proposed. The quantitative detection experiments on the depth parameter of defect in plate-like conductor has been implemented, which verifies that the studied method is effective.Theoretical analysis and test results show that the developed ECT/EMAT composite detection prototype system has made full use of advantages of both methods. The surface and internal defects can be simultaneously detected in a single scan. And the capabilities to detect the deep defect and identify the defect parameters have significantly improved. The research results in this paper have great application value for non-destructive testing and evaluation of important equipment.