Fault Recognition Research Of Mine Hoister Wire Ropes Based On Wavelet Transform

Wire rope has been widely used in mine production, metallurgy, traffic transportation, construction and tourist industry because of its light weight, good elasticity, high extension strength and fatigue strength. It plays a vital role especially in mine production. As a load-bearing component, it will inevitably involve the abnormal phenomena such as corrosion, localized fracture of wires and abrasion. These phenomena will gradually accumulate, eventually lead to the fracture of whole wire rope and severely influence the safety production of coal mine. Thus, fault recognition research of wire rope is of great significance for protecting personal and property safety in coal mine.This paper analysis and generalize the basic structure, classification and fault detection’s standard of wire ropes. After analyzing the structure of wire ropes, leakage magnetic field method is adopted to carry out the fault detection and appropriate excitation intensity is selected.This paper sets up the equivalent point dipole model and equivalent band dipole model to implement the numeric simulation, analyze how each parameter will influence the wire rope fault detection and make a comparison between those two models. In the meantime, finite element analysis is introduced by the software ‘ANSYS’ to get the magnetic field distribution regularities near the faults. Besides, we compare the two-dimensional magnetic field lines, magnetic vector and distribution regularities of magnetic induction intensity of broken wire rope with the normal wire rope and obtain the feature waveform of wire rope’s fault.We analyze the noise signal that could be involved into the wire rope fault detection and get the feature of these noise signals. By adopting the algorithm of multiple band-stop filter that based on the traditional band-stop filter, the strand noise of wire rope can be eliminated.Wavelet transform is introduced to carry out the preprocessing of wire rope fault data. Due to the frequency aliasing phenomena and bad translation invariance of traditional waveform, dual tree complex wavelet is selected to replace the traditional wavelet to implement preprocessing and analysis process. In the meantime, a comparison is made to distinguish the performance of frequency aliasing and translation invariance between traditional wavelet and dual tree complex wavelet. Besides, singularity analysis and multiple scale analysis are adopted to carry out the fault recognition of wire rope fault simulation signal.In order to verify the feasibility of these algorithms, a test bench is set up and signal acquisition software is made through ‘Lab VIEW’ to acquire wire rope fault data. Multiple band-stop filter, dual tree complex wavelet transform, singularity analysis and 120~128 scale analysis are adopted to implement the preprocessing and fault recognition process. Different numbers and types of broken wires(including internal and external broken wires) are analyzed. In the fault recognition process, an unexpected slight abrasion is detected which proves algorithms in this thesis is accurate and can be put into practical application.