| As a common key component of industrial production,steel wire rope has a large loadbearing safety factor,safe and reliable use,and has a wide range of applications in lifting,bearing and traction fields.Because its working environment is mostly in the open air,it is exposed to various stresses for a long time,and there are many kinds of damages such as wire rope wear,corrosion and broken wires.These injuries reach a certain level,ranging from affecting work safety,and threatening human life and health.The safe use of steel wire ropes is very important for national safety in production.Therefore,the research on the detection of wire rope damage and the development of an efficient and accurate wire rope damage detection device are of great significance to the safety of wire rope users and to promote the development of my country’s economy.This paper first introduces the common types and characteristics of wire rope damage,compares common wire rope damage detection methods,and conducts in-depth research on electromagnetic detection methods.By studying the principle of wire rope damage electromagnetic detection,a wire rope damage electromagnetic detection sensor is designed.Permanent magnet is selected as the excitation source,Hall element is used as the magnetic leakage detection element.The sensor adopts a multi-loop magnetic conduction method,and a magnetization device is added to collect the magnetic leakage.In order to ensure the accuracy of quantitative identification of wire rope damage,the original signal collected by the wire rope damage sensor was researched on denoising,and finally the wavelet soft threshold method was selected to denoise the original signal.The wavelet threshold denoising experiment was performed on the collected original signal in MATLAB,and the original signal was denoised by wavelet soft and hard threshold respectively.The signal-to-noise ratio and root mean square error of the signal after denoising were compared,and the wavelet soft threshold was found,the denoising effect is better,so the wavelet soft threshold method is selected to denoise the original signal of this damage detection experiment.According to the characteristics of the wire rope damage signal,the signal peak-to-peak value,wave width,area under the wave and wavelet energy are selected as the eigenvalues,and the eigenvalues are extracted and normalized.In order to realize the quantitative identification of wire rope damage,the method of quantitative identification of wire rope damage was studied.Research the classification methods of BP neural network,RBF neural network and support vector machine,and establish corresponding quantitative identification models of steel rope damage respectively.The experiment collected 30 groups of small-sample wire rope damage signal data,using the same test set and training set,to perform small-sample damage experimental identification tests on these three quantitative identification models,providing theoretical support for subsequent quantitative identification experiments.An experimental platform for wire rope damage detection was built,different types of wire rope damage samples were made,and wire rope damage detection experiments were carried out.After processing the collected LF damage signal and LMA type damage signal,the BP neural network model,RBF neural network model and support vector machine classification model of steel rope damage recognition were carried out quantitative recognition accuracy test experiments.The experimental results show that for LF-type damage recognition,the BP neural network model has the highest recognition accuracy when the number of hidden layer neurons is 5;for LMA-type damage recognition,the support vector machine classification model has the highest recognition accuracy and the best effect.The above research on quantitative identification of wire rope damage has laid the foundation for the application of quantitative identification of wire rope damage detection in practice. |
