Belt Machine Fault Diagnosis And Prediction Based On Support Vector Machine

With the increasing demand of high-speed transportation industry,in the continuous development of science and technology today,especially in the face of the rapid development of coal mining industry,in order to meet the needs of efficient transportation,belt conveyor transportation system needs to continue to develop toward efficiency,large and intelligent direction.But in this process,the internal structure of mechanical equipment gradually became more complex,and the components used in it also became more sophisticated.Because the whole machine itself is an organic and closely linked whole.In the engineering of machine operation,if any one part has a slight fault,it may bring the whole transportation system to a standstill.In this way,it will make all aspects of production work stagnation,will bring incalculable losses to enterprises.Therefore,reasonable use of intelligent control theory and related technology for belt machine timely fault diagnosis and prediction of the occurrence of the fault is particularly important.This paper will use support vector machine method,and it is applied to the belt machine fault prediction and diagnosis problem research,will use support vector machine classification and regression algorithm to build two models:one is the fault diagnosis model,the other is the fault prediction model.The main contents of this paper are as follows:(1)Since the identification accuracy of the belt machine fault diagnosis model itself is not very high,this paper adopts the following two methods to improve and optimize it.One is the use of principal component analysis(pca),the method can to extract fault characteristic value of the method 19 parameters can be transformed to 6 unrelated principal component index each other,thus exclude the characteristic indexes of correlation is not very tall,effectively avoid the errors produced by fuzzy factors for fault classification guide.The selection of superparameters of support vector machine will also affect the accuracy of fault diagnosis model.In order to eliminate this influence,the classical particle swarm optimization algorithm and the gray Wolf algorithm with better group search effect are initially selected in this paper,and the kernel parameters and penalty factors are optimized by using this method.In order to further improve the accuracy of the model and combine the advantages of the two algorithms,this paper finally adopts an improved gray Wolf algorithm based on particle swarm optimization to optimize the fault diagnosis and prediction model.Finally,the model is applied to the belt conveyor of ANHUI X Coal Mine Company.Through a lot of experimental analysis and comparison,the fault identification accuracy is finally improved to 97.22%,which further indicates that the fault diagnosis model proposed in this paper has better effect and meets the application standard of enterprises.(2)At the same time,this paper establishes a regression model to analyze the parameters of belt machine,and realizes the function of fault trend prediction through this model.Belt machine failure occurs,there will often be monitoring index changes with its occurrence,this paper chooses the change of motor temperature as the focus of prediction and analysis,and in the process of operation,belt machine equipment monitoring parameters are also affected by many cross factors,and the time series of index parameters is not regular.In order to solve the above problems,this paper uses the grey relational analysis method to extract the characteristic indexes.And using the five-fold crossover method and improved gray Wolf algorithm to optimize the prediction model,through comparison,the test results show that:the combined prediction model proposed in this paper has a good accuracy,compared with a single traditional model,it can make more accurate judgment of belt machine failure.Figure[58]Table[16]References[80]...