Research For Rolling Bearing Fault Diagnosis System Based On Nolinear

Rolling bearing is one of the most widely used components with high failure rate in rotating machines.According to the estimation,less than 10% bearing can operate normally as long as the ideal life length.Once the faults happened to the bearing,the consequences can be catastrophic,such as shut-down machine and casualties.With the development and application of science and technology,the traditional fault diagnosis methods such as time-domain characteristic parameter method and frequency-domain analysis method cannot meet the needs of modern production.As the result,the request for the corresponding ability of faults diagnosis is higher and higher.Early fault diagnosis is a key technique and research focus in this field.A weak fault usually refers to a weak or potential fault at an early stage,with characteristics of subclinical and weak feature information,which gives huge challenges for faults diagnosis.The nonlinear theory provides a new method and approach to overcome these difficulties.In this paper,based on the nonlinear theory,combined with the traditional fault diagnosis method,presents a research on early fault diagnosis of rolling bearing.In this paper,the vibration of the bearing is taken as the research object.The types and causes of faults are expounded and the formula for calculating the natural frequency and fault characteristic frequency is given.This paper introduces the definition and classification of nonlinear theory and the method to determine the chaotic state.A method based on fractal dimension is proposed to determine the critical threshold of Duffing oscillator from large period state to chaotic state.The Simulink simulation model is established in MATLAB by using nonlinear theory and Duffing oscillator.By comparing the results of the Duffing detector and the envelope spectrum analysis in the same rolling bearing fault data,the feasibility of Duffing oscillator detector in early fault diagnosis of rolling bearing is proved.A set of fault diagnosis system for rolling bearing based on virtual instrument technology is designed.Real time acquisition system and off-line fault diagnosis system are developed;it contains basic analysis module,time-frequency analysis module,characteristic frequency parameter calculation module,diagnosis module and database module.Through the experiment,the rolling bearing fault diagnosis system runs well,the interface is clean and the result of fault diagnosis can be obtained through the actual data and the result is accurate.So this system contains use value and the practical significance.