Wavelet De-noising Method Analysis of Pipeline Magnetic Flux Leakage In-line Inspection Based on Coefficient of Variation

Abstract

As an important national energy infrastructure, the oil and gas pipeline is known as the “lifeline project”. Magnetic flux leakage (MFL) testing is currently the most widely used inline inspection method for oil and gas steel pipelines, which can realize the identification, quantification and positioning of pipeline metal loss and other defects. MFL testing is equivalent testing. Eliminating noise in MFL signal plays a vital role in correctly extracting information from signal and realizing correct defect identification. In this paper, a wavelet de-noising method of MFL signal based on alternating coefficient is proposed. The method uses signal-to-noise ratio (SNR), root mean square error (RMSE), smoothness, for comprehensive evaluation. Combined with the time consumption, the appropriate wavelet denoising parameters are selected. The application of engineering inspection data shows that the method has a good application effect.