The invention belongs to the technical field of electromagnetic
nondestructive testing, and particularly relates to a decarburized
layer thickness detection method and device based on pulsed
eddy current and a medium. A probe comprising multiple coils and a self-calibration unit is placed on the surface of a workpiece to apply excitation, and positioning is completed; carrying out amplification and phase calculation on the received
signal, analyzing and optimizing a coil excitation
pickup sequence through sample piece pre-scanning, collecting a response
signal based on the optimized sequence, and extracting time-
frequency domain composite features; performing multi-frequency scanning on the workpiece by using the optimized sequence, extracting features of a scanning position, inputting the features into a pre-trained
deep learning model, and outputting thickness information; and judging whether a standard is met according to the thickness information. According to the method, by optimizing the coil excitation sequence and combining with the
deep learning model, the detection precision is improved, and nondestructive evaluation of the thickness of the decarburized layer can be effectively achieved.