Research On Local Defect Detection Method Of Mine Wire Rope Based On The Principle Of Magnetic

Wire rope is widely used in coal industry in China,and it is an important component of main and auxiliary shaft hoisting equipment in coal mine.Mine hoisting equipment is responsible for the transportation of coal and gangue,throwing equipment,carrying personnel and so on.Life-time use of wire rope will lead to potential safety hazards which are due to fatigue,such as wire breakage and wear.The safety of wire rope is directly related to the normal production of coal mine,the life safety of personnel and economic operation.At present,artificial visual check and regular involuntary conversion of steel wire rope are generally used to ensure safety in production.This method can not replace the steel wire rope which is seriously damaged due to various accidental factors in time,but also abandon the steel wire rope which is still usefull in many cases,resulting in huge waste.In the Notice of Effectively Strengthening the Safety Management of Mine Lifting Transportation,the State Administration of Safety Supervision clearly put forward that the safety inspection of lifting transportation system,especially lifting winch and wire rope,should be strengthened.Therefore,while continuing to upgrade the level of existing mine safety equipment,it is urgent to study on-line mine wire rope detection technology,prevent wire rope breakage scientifically and effectively,and reduce casualties.On the basis of summing up the nondestructive testing technology of wire rope at home and abroad,combining with the actual situation of mine wire rope detection,the structure and key theory of mine wire rope flaw detection sensor are deeply studied.The on-line testing platform for mine wire rope is built to effectively identify the broken wire damage of mine wire rope.(1)The principle of magnetic flaw detection for mining wire rope and the characteristics of defect magnetic signal are studied.The principle of magnetic flaw detection for mine steel wire rope is analyzed theoretically,and it is concluded that the flux leakage density is related to the defect cross-section area.Based on the principle of magnetic charge,the wire defect model of wire rope is theoretically studied.It is concluded that the magnetic signal of defect is related to the diameter of wire rope and the clearance of broken wires.The magnetic signal produced by the defect of the wire rope is analyzed numerically,and it is concluded that the magnetic signal of the defect mainly distributes in the axial direction of the wire rope.By studying the geometric parameters of defects,it can be concluded that the length of defects has less influence on the magnetic signal of defects,and the depth of defects and lift off values have greater influence on the magnetic signal of defects.(2)Three-dimensional numerical analysis and optimization of excitation structure of mine wire rope.Set up the three-dimensional mine wire rope excitation structure based on Ansoft.Finite element simulation and numerical analysis of twist wire rope model are studied.Compared with the traditional two-dimensional finite element static model,this model is more fit the actual working conditions of on-line detection of mine wire rope.The influence of each component of excitation structure on the magnetic signal of defect of mine wire rope is studied.The results show that the width of permanent magnet is the biggest factor to affect the magnetic signal of defect of mine wire rope.The mathematical relationship between the geometric parameters of excitation structure and the magnetic induction intensity in the magnetized wire rope is established.A permanent magnet ring radiation magnetization method is proposed,which can improve the uneven magnetization of mine wire rope better.Combining genetic algorithm with quadratic nonlinear programming function,the excitation structure is optimized and the optimal size of excitation structure model is established.(3)A multi-hall element array detection method based on magnetic concentrating principle is proposed,and a testing platform for mine steel wire rope detection is built.Eight differential Hall elements of model HW-302 B are selected as magnetic sensors to realize all-round leakage detection of wire rope with fewer detection elements.A ring magnetic concentrating structure is designed,which can enhance the defect magnetic signal of 10 times.The equivalent magnetic circuit method is studied to optimize the magnetic concentrating structure and improve the magnetic concentrating efficiency.A multi-Hall element array detection method is proposed to eliminate the influence of magnetic signal between strands of wire rope on defect signal.Structural Design the structure of test platform for mine wire rope.Through three-dimensional dynamic finite element analysis,the influence of wire rope running speed on defective magnetic signal is studied,and finally the model selection of motor is completed.The structure of flaw detector,detection circuit and data acquisition module are designed.Finally,an online testing platform for mine wire rope is built,and different degrees of damage are detected and analyzed.(4)A preprocessing method of damage signal of mine wire rope based on wavelet packet-sparse representation is proposed.Aiming at the collected damage signal of mining wire rope,a noise reduction method based on wavelet packet-sparse representation is proposed,which is suitable for the damage signal of mine wire rope.The original signal is decomposed into three layers of wavelet packet,and the wavelet packet coefficients are sparsely represented by MP and OMP algorithms.The signal-to-noise ratio of reconstructed signal is much higher than that of wavelet threshold shrinkage method,median filter method and singular value difference spectrum method,which has a remarkable effect on noise reduction of mine wire rope damage signal.(5)A PCA-based PSO-SVM model for quantitative identification of broken wire signal of mine wire rope is proposed.Seven global eigenvalues and wavelet packet energy entropy details of broken wire damage signal are extracted as effective eigenvalues.The eigenvalue dimensionality reduction of eight eigenvalues is used as the input of support vector machine.The support vector machine is designed and trained.A PSO-SVM classification model based on PCA is proposed.Compared with the recognition efficiency before and after optimization,the recognition rate of SVM is 94.73%,and the quantitative recognition accuracy is higher.