| This paper is supported by the State key R & D Program((No.2016YFF0203000),the National Natural Science Foundation(No.51675258,No.51261024)and the Science and Technology Research Project(No.GJJ150699)of Jiangxi Education Department.The adaptive Fourier decomposition algorithm is introduced into the feature extraction of magneto acoustic emission signals in the early fatigue state of ferromagnetic metal materials.An analysis method of characteristic parameters of magneto acoustic emission signals based on adaptive Fourier decomposition is proposed.This paper mainly discusses the following aspects:In the first chapter,the significance of this topic is introduced,and the research status of magneto acoustic emission detection technology and adaptive Fourier decomposition(Adaptive Fourier Decomposition,AFD)at home and abroad are summarized.Aiming at the deficiency of extracting characteristic parameters of magneto acoustic emission signals,the research contents and innovations of this paper are given.In the second chapter,we first introduce the Hardy Space decomposition and rational orthogonal system.Then,the theory of single component function is systematically discussed from three aspects: single component function,single component function representation of signal and definition of inner and outer functions.On this basis,the AFD,of spatial function is derived.In this paper,the design principle of three typical algorithms of AFD(Core AFD,Cyclic AFD and Unwinding AFD)is discussed.At the same time,the three typical algorithms are compared and analyzed.The simulation results show that the principle of selecting one-component function by relative Core AFD,Cyclic AFD algorithm,Unwinding AFD algorithm is term-by-term optimal,so that the partial sum can approximate the original function as much as possible in terms of energy,so that the original function can be approximated to the maximum in terms of energy.The energy difference between the reconstructed signal obtained by AFD algorithm and the original signal is smaller and smaller.In addition,Unwinding AFD algorithm has obvious advantages over Core AFD,Cyclic AFD refueling in terms of convergence speed.In the third chapter,aiming at the shortage of the existing magneto acousticemission signals which are easily disturbed by the noise and the unique advantages of AFD,a method of extracting the characteristic parameters of the magneto acoustic emission signals under static load based on adaptive Fourier decomposition is proposed,in which the proposed method is used to extract the characteristic parameters of the magneto acoustic emission signals.Firstly,several single-component functions are obtained from the original signal collected by AFD.The energy difference of each single-component function is compared with that of the original signal,and each single-component function with the smallest energy difference is reconstructed to eliminate independent interference components.Then the MAE feature parameters are extracted based on the reconstructed signal.At the same time,with the traditional magneto acoustic emission signal characteristic parameters The experimental results show that the proposed method can obviously improve the quality of the original signal and the signal-to-noise ratio(SNR)of the analytical signal,according to the(SNR)index of signal-to-noise ratio(SNR)of the signal-to-noise ratio(SNR).In the analysis of the interfered magneto acoustic emission signal,the method of extracting the characteristic parameters of the original magneto acoustic emission signal directly does not get the ideal result.However,the proposed method based on adaptive Fourier decomposition can avoid the influence of external interference,and can still reflect the variation rule of each characteristic parameter curve under static load and tensile condition well,therefore,the proposed method can be used to extract the characteristic parameters of magneto acoustic emission signals under static load and tensile condition.The proposed method is obviously superior to the traditional Method of characteristic parameters of Magneto acoustic Emission signal.In the fourth chapter,adaptive Fourier decomposition is applied to the feature extraction of the magnetic sound emission in the fatigue state,and the mapping relation between the characteristic parameters and the cycle time of the magnetic sound emission signal based on the self-adaptive Fourier decomposition is studied in the low-cycle and high-cycle fatigue state.The experimental results show that the characteristic feature extraction of the magnetic acoustic emission signal under the condition of fatigue can not reflect the variation of the characteristic parameter curve under the fatigue state.The method for extracting the magnetic sound emission signal based on the AFD overcomes the defect of the traditional magnetic acoustic emission signal characteristic extraction,The influence of external interference on the magnetoacoustic emission signal is greatly reduced,and the mapping relationship between each characteristic parameter and the cycle is better reflected.In the fifth chapter,the characteristic parameters of the existing magneto acoustic emission signals are all dimensional parameters,and the values of the characteristic parameters of these dimensional magneto acoustic emission signals are often changed due to the change of load,rotational speed,tension and other conditions,so it is very difficult to distinguish between them in practice.The improvement method is to introduce the characteristic parameters of the magneto acoustic emission signal with one dimension.They are not sensitive to the change of the amplitude and frequency of the MAE signal,that is to say,it has little relation with the working conditions of the machine,but is sensitive to the change of tensile stress.In this paper,four dimensionless characteristic parameters of magneto acoustic emission signal,namely waveform index,pulse index,peak value index and kurtosis index,are presented.Their definitions and algorithms are given..On this basis,combined with the dimensionless characteristic parameters of AFD and magneto acoustic emission signals,a method of extracting dimensionless characteristic parameters of magneto acoustic emission signals based on adaptive Fourier decomposition is proposed.The method is applied to the feature extraction of magneto acoustic emission signals under static load tension and fatigue conditions,and is compared with the method based on adaptive Fourier decomposition for extracting the dimensional characteristic parameters of magneto acoustic emission signals.The experimental results show that the dimensionless characteristic parameters of magneto acoustic emission signals are more sensitive than those based on adaptive Fourier decomposition.The proposed method is clear.The obvious advantage can reflect the mapping relationship between each dimensionless characteristic parameter and cycle.The results show that by comparing the dimensionless and dimensionless characteristic parameters,it is found that the proposed dimensionless characteristic parameters have more obvious relation to the micro-structure and stress state of the materials,and have higher sensitivity.In the sixth chapter,the research results obtained in this paper are analyzed and summarized comprehensively,and some problems worthy of further study are given. |
PDF Full Text Request