Research On Extraction And Process Of Fault Signal In Vertical Roller Mill

The condition monitoring and fault diagnosis of Important mechanicalequipments is of great significance to keep these equipments running smoothly andwithout accident. Fault diagnosis system consist of three parts,data acquisition,feature extraction and fault decision. The feature extraction of fault signal plays aimportant part in the fault diagnosis procedure.Vertical roller mill is a kind of large equipment used in the in cement industry.It’s used to crack and dry Cement. Vertical roller mill is always running in theenvironment of heavy load, high pressure and high temperature, the vibrationsignal of the vertical roller mill is random, multidimensional, time-varying, andnonlinear. Currently, research on fault diagnosis of vertical roller mill is rarely seen.This paper is mainly focused on the fault diagnosis of vertical roller mill. First,finite element analysis is done to get to know the structural characteristics ofvertical roller mill. Classic signal analysis method is not suitable in such circumstance.In this paper, we propose suitable analysis method based on short time Fouriertransformation and wavelet package transformation which has good performance todistinguish between normal signal and fault signal. A fault diagnosis system forvertical roller mill is developed on Java Web technology.The design, coding and testof the diagnosis system is done. Test shows the system has great performance.The main work of this paper is as follow:(1) Finite element analysis is done to get to know the structural characteristicsof vertical roller mill. By simplifying the structure of the vertical roller mill, weobtained its modal vibration mode using ANYSIS.Based on finite Element Analysiswe did the simulation of random vibration.(2) Based on short-time Fourier transform and wavelet packet transform, twofeature extraction method applicable for the vertical roller mill are proposed. Tests using MATLAB simulation show these methods has good performance.(3) A fault diagnosis system for the vertical roller mill and its feature extractionmodule is developed using the method proposed before. Test of this system is doneand shows good performance.