Research On Fault Diagnosis And Prediction Of Belt Conveyor Based On Support Vector Machine

With the improvement of science and technology,the belt conveyor equipment is developed to be more intelligent,larger scale and higher speed,which greatly promotes the efficiency of trasport.However,the improvement and innovation of equipment have brought some problems.The structure of equipment becomes more complex and the connection of each component is more closely,which makes the minor faults of single component may lead to the paralysis of the whole transportation line,and then bring great losses to enterprises.Therefore,the timely fault diagnosis and operation status prediction of belt equipment is necessary to ensure the safe transportation.This paper applies the Support Vector Machine(SVM)in the fault diagnosis of the belt conveyor,where the SVM classification algorithm and regression algorithm are used to construct the fault diagnosis model and the fault trend prediction model.And the main work of this paper are as follows.1.Aiming at the problem that model of belt conveyor has lower accuracy for fault identification,this paper uses Principal Component Analysis(PCA)and improved Grey Wolf Optimizer(GWO)to optimize the model.Firstly,the Principal Component Analysis is used to extract feature vaules,by which 19 parameter indicators are transformed into 5 mutually unrelated principal component vectors.In this way,the low-correlation feature indicators are eliminated,and then the misleading fault classification caused by fuzzy factors can be avoided.Secoudly,in order to eliminate the influence of SVM hyperparameter selection on the classification model,the Grey Wolf Optimizer with better group search effect is uesd to optimize the kernel parameters and penalty factors.Especially,this paper adopts a differential evolution improved Grey Wolf Optimizer which better adapt to the model to enhance the accuracy of the model.Finally,the model is implemented in the example of the belt unit of X Coal Mine in Shandong Province.And through the analysis and comparison of the experiment,the recognition accuracy of the fault is improved from 77.78% to 97.22%.In summary,the combined classification model has better fault diagnosis and meets the requirements of enterprise application.2.In this paper,a regression analysis model is established for the parameters of the belt conveyor to predict its failure trend.The occurrence of belt conveyor failure symptoms is often accompanied by changes in monitoring indicators,and When the belt conveyor is in the state of running,the parameter variation of the equipment isaffected by many factors,and the time series index of the parameters is generally irregular.In order to solve this problems,this paper uses grey correlation analysis to select the eigenvalue indicators,where an accumulation operation on the time series is performed to improve the regularity of the sequence itself.Finally,the validity of the prediction model is verified by the monitoring data of belt conveyor in X Coal Mine,Shandong Province.And the experiment shows that the combined prediction model proposed in this paper has better prediction accuracy than the single method,which means that the combined model can accurately judge the failure trend of the belt conveyor.