Improved Reconstruction of Magnetic Flux Leakage Signal From Wire Rope Damage Using Novel Smooth Function and Feature Peak Compensation

In magnetic flux leakage (MFL) detection of wire ropes (WRs), strand noise cannot be avoided, which seriously interferes with extracting local damage feature peaks. This article innovatively constructs a phenomenological model of axial WR detection signal, designs a novel smooth function, and proposes a new method for high-fidelity reconstruction of damage feature peaks. First, a phenomenological model of the axial signal is established to more clearly characterize the waveform difference between strand noise and damage signal. Second, a novel smooth function is constructed, which increases the rate of change of the threshold function and makes it approach the original value smoothly at the threshold. Then, a method of signal denoising and feature peak compensation (FPC) is proposed, which can remove the strand noise and reconstruct the damage feature peak with high fidelity. Finally, compared with nine different methods, the proposed method performs well on signal-to-noise ratio (SNR), peak-to-peak (PTP) ratio, false positive rate, coverage rate, and other indicators, which proves that the proposed method has the highest signal denoising ability and the highest fidelity of feature peak reconstruction.