Fault Diagnosis Of Mechanical Transmission Device Based On Hidden Markov Model

Mechanical transmission is an indispensable part of industrial products.Because it is a complex system,whether it is processing or assembly,the required accuracy is relatively high,coupled with it often working in state of sustained high load,failure is also normal.After a fault occurs,abnormal vibrations and noise may be generated during operation of the transmission,which may even result in the inability of the entire device.Therefore,it is necessary to study the transmission fault diagnosis.During use of transmission,if fault can be discovered early and it can be diagnosed,the use or maintenance plan can be reasonably formulated to prevent the occurrence of accident during production.The fault diagnosis technology based on Hidden Markov Model(HMM)is adopted in this paper.The method has stronger classification ability,less demand on sample and faster calculation speed,and is more suitable for analyzing the nonlinear vibration signal.Main task of this paper is to use HMM for transmission fault diagnosis,including following aspects:1.The research significance of transmission fault diagnosis were discussed.The common methods of fault diagnosis technology and the application of HMM in fault diagnosis were summarized.The common fault types and causes of transmission were analyzed,and various fault diagnosis techniques were compared.2.The role of Fast Fourier Transformation(FFT)and AR model power spectrum estimation in transmission health monitoring is analyzed and verified.Feature extraction based on Ensemble Empirical Mode Decomposition(EEMD)was used,and it is applied in DHMM.Functions based on C++ programming language was compiled.3.The basic theory of HMM is discussed,and the implementation of fault diagnosis for Discrete HMM(HMM)is mainly studied.Functions based on C++ programming language was compiled.Finally,the HMM-based fault diagnosis idea and flow were given.4.State monitoring and fault diagnosis software was developed based on Qt and C++ programming languages.The software is easy to use,and it has a friendly and intuitive graphical interface.It can quickly and accurately complete monitoring and diagnostic work on the experimental platform built in this paper.