Research On Recognition Method Of Broken Wire In Wire Rope

Steel wire rope plays as an important role in hoisting system as it a significant part of mine hoist.For the reason that steel wire rope always works under high load,failures like abrasion,shape change and even the broken wire,which will be a potential menace for mine safety.As a typical form of wire rope damage,broken wire has always been concerned in wire rope condition monitoring.Hence,research on the detection on wire breakage of steel wire rope is of great significance to ensure the security of hoisting system.In this research,theory,simulation and experimental study of broken wire detection are conducted based on magnetic flux leakage(MFL).The main research contents are as listed as follows.Firstly,the structure and damage form of wire rope are analyzed,and the overall scheme of magnetic flux leakage detection is determined for wire damage.By equivalent band dipole model,the influence of damage degree and lift-off value on leakage magnetic field is simulated and analyzed.It is found that the peak value of the radial component increases with the damage degree,and the axial and radial components of the magnetic flux leakage signal decrease with the increase of lift-off value.Secondly,an experimental detection system of broken wire is designed.The dynamic experiment table is designed and the adjustable frame and motor control system are used to simulate the running condition of the wire rope.A damage magnetic flux leakage signal detection device is designed,and the axial,radial and circumferential magnetic flux leakage signals are collected simultaneously by the annular hall element array.The signal acquisition system on host computer based on LabVIEW is developed for high-speed data acquisition and storage of 24 channel magnetic flux leakage signals.The effect of damage degree of broken wire on magnetic flux leakage signal is analyzed experimentally,which verified the validity of the wire rope testing system.Thirdly,the application of lifting wavelet transform method in the denoising of magnetic flux leakage signals caused by broken wire breakage in wire rope is studied.The axial and radial magnetic flux leakage signals are denoised by selecting lifting scheme,setting signal threshold and combining soft threshold processing scheme.The feature value extraction of the de-noised signal was carried out,and the feature value extraction methods of peak to peak value,bandwidth,area under waveform and wavelet energy were studied and characteristic value samples of wire rope damage identification with different number of broken wires are constructed.Finally,the quantitative recognition model of broken wire based on BP neural network is built.Support vector machine(SVM)model and majorization for SVM based on particle swarm optimization(PSO)algorithm are studied.The quantitative recognition method of broken wire based on PSO-SVM has the highest recognition accuracy.