Study On Failure Mechanism And Diagnosis Method Of Pipe Saw

The pipe row saw is a technological process of sawing seamless steel pipe to a certain length with a saw blade.In this system,there are many equipments,many monitoring points and data sampling signals,which objectively result in a series of problems,such as high failure rate,long maintenance period,loss of shutdown,etc.production stagnation and difficulty of maintenance and detection are increased due to the failure.In order to better identify and determine the location of equipment failure,only all aspects of testing can be carried out,resulting in increased workload of maintenance personnel and reduced work efficiency.Therefore,it is of great significance to diagnose the fault type of pipe saw,to ensure the reliable operation of mechanical equipment,and to avoid huge economic losses caused by enterprises.First of all,it introduces the sawing process of the pipe saw and the main machine fault,hydraulic system fault and auxiliary machine fault in the equipment.Find out the main saw blade failure and main machine clamping seat failure which have great influence on sawing technology,and analyze the reasons and mechanism of main saw blade wear failure,electromagnetic reversing valve card and proportional reducing valve card failure.The common saw blade wear failure,pressure valve card failure at the inlet of positioning side and displacement valve card failure at the inlet of clamping side were selected.Based on the data monitored by the upper computer,the data of these 10 fault conditions are extracted,and the eight time-domain features which can best represent the operation state are selected from the time-domain feature parameters.Because of the redundancy and correlation between these features,dimension reduction method is used to reduce the operation of SVM.Then based on principal component analysis(PCA)and kernel principal component analysis(KPCA)theory.Through the matlab program,eight features are transformed into three features which can best reflect the fault characteristics.Then,PCA-SVM and KPCA-SVM fault diagnosis models are established by combining them with support vector machine.The accuracy of PCA-SVM fault classification is 85%,while that of KPCA-SVM is92%.It is verified that KPCA-SVM has better classification accuracy,and the classification result will be used as the input of fault diagnosis interface.Finally,the data transfer between WinCC and MATLAB is realized based on OPCcommunication technology.In the instruction window of MATLAB,the program needed for KPCA-SVM classification result is written,and then the data of classification result is transmitted to WinCC software.The specific fault type and location can be displayed in the interface of fault diagnosis system for main saw blade and clamping base of main machine of pipe row saw.Therefore,the method used in this paper has certain practicability for the fault diagnosis of the pipe saw system,and provides a new idea for the fault diagnosis of other process systems.