Research On Defect Detection Of Ferromagnetic Materials Based On Weak Magnetic Method

Ferromagnetic materials have a wide range of applications in electronics,information technology,energy,and other fields due to their excellent magnetic properties,such as high magnetic permeability,low magnetic resistance,and easy magnetization.With the increasing use of various ferromagnetic devices,which often operate in high voltage,high load,and high temperature environments,the performance of device materials is more susceptible to damage,which seriously affects the safe operation of the equipment and poses a threat to the surrounding environment and personnel safety.Therefore,defect detection of ferromagnetic materials is particularly important.Weak magnetic detection technology is an important way of non-destructive testing of defects and has accurate identification capabilities for macroscopic defects and early hidden damage of materials.This thesis studies the problems of weak magnetic detection in defect detection of ferromagnetic materials under actual conditions,and the specific research contents are as follows:1.Review and analyze the literature in this research field at home and abroad.Based on some problems existing in the current research,this thesis proposes the research ideas.2.Research on weak magnetic detection mechanism.In the quantitative relationship between stress,defects,and weak magnetic signals,this thesis establishes a magnetic defect on weak magnetic signals,and preliminarily explains the basic experimental phenomena and laws of weak magnetic detection.The establishment of the weak magnetic signal database of ferromagnetic materials with defects has been completed.3.Research on signal preprocessing.In response to the problem of mixed noise in weak magnetic signals collected in practical applications,this thesis analyzes the causes and characteristics of noise in weak magnetic detection,and finds that weak magnetic signals are nonlinear blind mixed signals with noise interference.It proposes using postnonlinear blind source separation algorithm for blind separation experiments of weak magnetic signals.Experimental verification shows that the post-nonlinear blind separation algorithm based on kernel functions can effectively separate weak magnetic blind mixed signals,accurately retain defect features,and provide an effective method for non-destructive testing of ferromagnetic materials.4.Selection and optimization of classifiers.In the quantitative estimation of defects based on weak magnetic signals,genetic algorithm and BP neural network are combined to calculate and select multiple feature combinations using weak magnetic signals collected by magnetic flux gate magnetometers.A mathematical model based on weak magnetic detection features and ferromagnetic defect categories has been established to achieve multi-defect classification recognition and rectangular defect 1-5mm width and depth grading prediction,providing a methodological basis for quantitative research on defect detection of ferromagnetic materials.In summary,this thesis deeply explores the problem of ferromagnetic material defect detection,proposes multiple technical methods,and comprehensively considers signal acquisition,signal processing,and data analysis.Through in-depth analysis of the quantitative grading of ferromagnetic material defects,the defect detection and quantitative grading model proposed in this thesis accurately inverts the defect type of the ferromagnetic specimen,providing important inspiration and guidance for future research and engineering applications.