Support Vector Machines Based On Wavelet Packet And Optimization Research On Fault Diagnosis Of Rolling Bearing

In the equipment manufacturing industry,rolling bearings,as an important supporting component of machinery,have the title of mechanical "joint" and play a key role in the development of industrial manufacturing.Their operating status is directly related to the overall operating safety of the equipment.The rolling bearing vibration signal contains various important information.The vibration signal is analyzed-filtered-extracted through reasonable signal processing technology,and then the extracted characteristic information is identified,which can be used for fault diagnosis and health monitoring of the bearing.Good function.This article starts with the vibration signal of rolling bearings,and conducts fault diagnosis and system development of rolling bearings.First,the background meaning and vibration principle of rolling bearing fault diagnosis are explained,and the corresponding characteristic frequency when the rolling bearing is damaged is given,and it is used as a sign to detect whether the rolling bearing is faulty.Then,the method of extracting the fault characteristics is discussed:Introduce the wavelet packet algorithm and the CEEMDAN algorithm to decompose the bearing vibration signal,and extract the fault features based on the energy spectrum and the correlation coefficient respectively,and finally reorganize the extracted fault features and compare them based on the Hilbert envelope spectrum algorithm For the two extraction effects,it is determined that the fault feature extraction effect based on wavelet packet algorithm is better.Secondly,the pattern recognition algorithm of SVM is studied,and the GS and PSO algorithms are used to optimize the parameters that affect the classification effect of the classifier in the support vector machine;combined with the existing experimental platform to collect the rolling bearing vibration data,use the optimized support vector machine Train the fault feature vector,compare the classification effects of different feature extraction algorithms and pattern recognition algorithms,and select the best feature extraction-pattern recognition combined algorithm for system design.Finally,the system is developed through the MATLAB GUI graphical user interface,and the overall framework of the system is designed.It realizes the model management of rolling bearings,time-frequency analysis of vibration signals in time domain,frequency domain and wavelet packet changes,intelligent diagnosis of faulty rolling bearings and other functions,combined with faulty rolling bearings under actual working conditions,to carry out the developed system The verification results confirmed that the system works well and plays an important role in the diagnosis of rolling bearing faults.