Analysis And Research Of Magnetic Flux Leakage Signal In Sucker Rod Based On Wavelet Analysis

Sucker rod is one of the important parts in rod pumping equipment with a large need. The most common accident is fracture with rod body and screw thread.The reason is more complex operating environment during its use, sucker rod produce a variety of defects under the action of alternating stress which further causes rod downhole fracture.That directly affects the safty of sucker rods and then causes significant economic losses. Therefore, quantitative nondestructive testing for sucker rod in use can greatly reduce rod fracture accidents, decrease oil production costs, and also has practical significance.This paper studies the parameters quantitative analysis of sucker rod defects in magnetic flux leakage testing. A new method was proposed for this probem:After the establishment of a signal template library, signal extraction after de-noising matching with the signal of template in the library, finding the equivalent template, finally to realize the parameters quantitative analysis.Using ANSYS finite element analysis software to analyse the relationship between defects and leakage magnetic field for its great speed and accuracy in calculation.By comparing to fact detection signals, it proved this method can be used in building the defect signal template library for its advantage.Then introduced the theory of wavelet analysis and wavelet denoising, hard and soft threshold denoising method were proposed, and selected a soft-threshold method to the actual signal preprocessing of sucker rod defects.In the signal extraction, introducted the threshold method to select the feature space of defect signal, finally completed the extraction of defect signal.The method of signal recognition is proposed based on cross-correlation theory. In this method, by the waveform of defects signal matching with various types of defects in the signal template, taking the template type corresponding to the largest correlation coefficient as the defect type corresponding to the defect signal. At last realized the inversion process from a given defect leakage field to the quantitative analysis of defects.