Study On The Prediction Of The Wear Status Degraded Of Pick Based On Principal Component Analysis

As the main cutting part of mining equipment,pick is widely used and important.It is urgent to study the wear degree,wear degradation law and the degradation state of shearer pick.How to realize the evaluation and prediction of the degradation state of the pick is an urgent problem to be solved at present.According to the quality of the pick and the size of the alloy head of the pick,the wear degree of the pick is divided into 6 kinds,which are defined as new teeth,slight wear,moderate wear,medium wear,and large wear,respectively,through the construction of the experimental platform,and the multi-signal acquisition during the process of pick wear.Severe wear,failure,pick-up.The cutting signal is collected according to the wear degree of the pick,and the influence of the noise on the signal acquisition is considered,and the noise reduction processing is done by using the wavelet dedryness.Using wavelet packet decomposition theory,the acceleration energy of vibration signal 50 ~ 100 kHz and acoustic emission signal 12.5~50kHz frequency band are collected and taken as sample data.In this paper,a grey prediction model based on degraded data and a Gamma model based on Bayesian parameter updating are established to predict the degradation of pick wear.The results show that the relative error of grey prediction under vibration signal is only 0.45%,through the application of data sample,the relative error of grey prediction is only 0.45% under vibration signal.The relative error of Gamma model is 0.57%,and the relative error of Gamma model is0.22% under Bayesian updating.The relative error of gray prediction under AE signal is 3.92%,the relative error of gamma model is 2.53%,and the relative error of Gamma model under Bayesian update is 2.19%.The prediction accuracy of the three models is very high.Well,the prediction error of Gamma model under Bayesian update is minimal.Based on the principal component analysis(PCA)theory,the problem of estimating the degradation state of the pick is studied.Based on the sum of acceleration energy in the frequency band of vibration signal,the sum of acceleration energy in the frequency band of acoustic emission signal,the current signal and the wear amount,four parameter changes are presented.Through the normalization of parameters,the weight of principal parameter is determined,the comprehensive score model of principal component analysis is established,and the evaluation curve of the degradation state of the pick is obtained.According to the curve,the wear state of the pick is gradually degraded,which is in accordance with the law of the degradation of the pick.Among them,new teeth,slight wear,moderate wear,low degradation rate in the early stage ofmedium-large wear,severe wear in the late stage of wear,and accelerated degradation rate in the wear failure period.The paper has 22 diagrams,22 tables and 197 references.