Design And Research Of Intelligent Detection System For Wire Rope Damage In Underground Mines

Steel wire ropes are made of high-strength steel wire and are characterized by high tensile strength,light self weight,and good flexibility.They are used in traction,pulling,and other aspects in mines.Steel wire rope,as a key load-bearing component,inevitably experiences defects such as wire breakage,wear,and deformation during long-term operation in the harsh environment of mines,posing a potential threat to production safety.Therefore,timely,efficient,and accurate detection of steel wire ropes has important economic and social significance for production enterprises.In actual working conditions,workers not only need to detect damage to the steel wire rope through magnetic flux leakage signals,but also need to observe the morphology of the damaged part of the steel wire rope.This article is based on the non-destructive testing technology of steel wire ropes and conducts relevant design and method research on the intelligent detection system for underground steel wire rope damage.The main research content is as follows:(1)In order to explore the detection of steel wire rope surface damage based on magnetic flux leakage signals and low illumination images,this thesis designs a steel wire rope damage signal acquisition experimental platform based on magnetic flux leakage detector and machine vision,which is used to collect steel wire rope magnetic flux leakage signals and surface image data.For the test-bed,the overall structure design,device selection,and statics analysis of the slide rail are carried out to ensure the normal shooting of the industrial camera.At the same time,a detailed introduction was also given to the selection and planning of steel wire ropes,data collection plans,and experimental steps in the experimental plan.(2)In response to the problem of wire rope magnetic flux leakage signal damage detection,classical methods overly rely on the user’s prior knowledge,accuracy,and generalization performance,which are unstable and difficult to locate damage.Therefore,this article proposes a method for detecting and identifying magnetic flux leakage signal damage in steel wire ropes based on the short-term kurtosis method.Compared with the threshold discrimination method based on wavelet transform,the proposed method can effectively detect and locate wire rope damage.At the same time,in the signal preprocessing process,this article uses the adaptive shift average algorithm to denoise the signal,and compares it with various denoising methods.The results show that this method can effectively denoise the magnetic flux leakage signal of steel wire ropes.(3)In response to the problem of low illumination image damage detection for steel wire ropes,the images collected by industrial cameras have problems such as uneven lighting,low contrast,and blurry details due to factors such as the underground environment of mines,making the detection and recognition of steel wire ropes more difficult.Therefore,this article proposes a low illumination image damage detection and recognition method for steel wire ropes based on Conv Ne Xt.We use logarithmic transformation and homomorphic filtering to address the issues of uneven lighting and low contrast,and use guided filtering to fuse the images processed by the two methods.Then,the fused image is divided into high-frequency and low-frequency images through non downsampling contour wave transformation to suppress noise in the lowfrequency images,enhance details in the high-frequency images,and finally perform inverse transformation to achieve image enhancement.This article uses Conv Ne Xt network to classify and recognize images,achieving the detection and recognition of low illumination damage images of steel wire ropes.(4)In order to integrate the previously proposed method for detecting and identifying magnetic flux leakage signal damage in steel wire ropes based on the shortterm kurtosis method and the method for detecting and identifying low illumination image damage in steel wire ropes based on Conv Ne Xt,this article uses Python to write the underlying code of the algorithm,and Py Qt5 to create a GUI interface,resulting in an intelligent detection system for underground steel wire rope damage in mines based on Python.The system can automatically achieve real-time detection of magnetic flux leakage signals and images while collecting and storing magnetic flux leakage signals and surface images of steel wire ropes.Moreover,the system can provide real-time feedback on the number and location of damages.Users can also observe the original and enhanced images of each damage location through image diagnosis,assisting in understanding the current morphology status of the steel wire rope.This thesis has 77 figures,11 tables,and 98 references.