Research On The Development And Application Of Weak Magnetic Detection System For Remanufacturing

As a new weak magnetic non-destructive testing method,metal magnetic memory technology has a good application prospect in evaluating the remanufacturing performance and service performance of ferromagnetic components.This method can detect and early diagnose the damage of workpiece structure.With the addition of excitation,the stress concentration in the ferromagnetic material is detected and evaluated;however,the traditional magnetic field measurement is mostly uniaxial measurement,the magnetic field information is not comprehensive enough,and the detection equipment is not portable enough,which limits the magnetic memory detection.Application scenario.Aiming at the above limitations,this paper develops a set of weak magnetic detection equipment based on embedded,and performs magnetic memory signal detection on tensile loading of samples.Quantitative research on defects by magnetic memory signal characteristic parameters further improves the weak magnetic field.Application of detection technology.This paper first studies the development of weak magnetic detection system.The hardware platform implementation system uses the OKMX6UL-C development board based on the i.MX6 UltraLite Cortex-A7 architecture processor as the core module for data processing and hardware control.The peripheral sensor probe is designed with HMC5883 L three-axis magnetoresistive sensor to realize the function of the system..The software platform uses Ubuntu system to build Qt development environment,and uses C language and C++ language to mix magnetic signal acquisition program,IIC communication program and data waveform display program to complete the easy-tooperate human-computer interaction interface based on Qt/Embedded development.Secondly,the calibration and operation debugging of the weak magnetic detection equipment is completed.The calibration system based on Helmholtz coil is built to test the static characteristics of the embedded detection equipment,and the calibration of the instrument is completed and the corresponding performance indicators are obtained.The measurement range of the system is determined within ±8Gs,and the resolution of the three axes is determined.The rates are 4.51,4.4,and 4.64 mGs.By detecting the samples of demagnetization and magnetization,it is determined that the detection equipment can correctly and effectively detect the characteristic values of the metal magnetic memory tangential signal and the normal signal,and compare with the data collected by the Russian TSC-2M-4 detector.The results are similar,further demonstrating that the testing equipment can correctly detect ferromagnetic workpieces.Finally,quantitative evaluation of magnetic memory signals based on PCA and GABP neural networks was completed.By making a sample with crack defects and performing tensile test,and using the self-developed weak magnetic detection system to detect the crack magnetic information of the defect,based on the analysis of the characteristics of the magnetic memory signal waveform,according to the data characteristics of the magnetic memory signal The feature index of the defect is extracted,and the combination of principal component analysis and genetic algorithm is used to optimize the neural network.The quantitative analysis of the defect signal not only removes the data correlation,reduces the sample dimension,but also simplifies the network structure and improves the structure.The global searchability of the algorithm can effectively reflect the degree of damage of the defect.