Rotating Machinery Fault Diagnosis Research Based On Axial Trajectory Combination Moment And Support Vector Machine

In the field of modern machinery manufacturing,rotating machinery manufacturing technology and fault diagnosis technology have developed rapidly with the development of science and technology and the requirements of industrial enterprises.Bearings and rotors are important components of rotating machinery equipment,and corresponding operational safety and reliability requirements are getting higher and higher.Failure of the rotor bearing components can also cause damage to other parts of the mechanical equipment,which can cause a chain reaction,sometimes it can damage the mechanical equipment,even it cause accidents and influence the safety of the equipment and the life of the operator,and affect the safety and economy of the rotating machinery directly.Therefore,the research on monitoring and fault diagnosis of rotating machinery has great economic and social significance.This paper studies the purification,identification,fault diagnosis system and diagnostic algorithm of the axis of the rotating machine,it's main contents are as follows:(1)According to the principle of rotating machinery failure,the trajectory library of the axis is established by using MATLAB,and the Gaussian signal is introduced to establish the noise map.Aiming at the existing noisy axis trajectory map,the wavelet denoising method is used for image denoising.By changing the wavelet denoising hard and soft threshold parameters,the axial trajectory denoising ability of the two methods is compared.The results show that the soft threshold wavelet denoising method has better denoising effect on the axial trajectory.(2)According to the axis trajectory after denoising,the image features are analyzed.The Hu invariant moment and the wavelet invariant moment are used as the features of the axis trajectory.By comparing the characteristics of the two invariant moments,the form of the combined moment is selected as the axis.The characteristics of the heart trajectory.The results show that the feature aggregation degree of the combined moment is better,and the recognition efficiency and accuracy of the axial trajectory are obviously improved.(3)For the wavelet invariant moment,data redundancy is not conducive to the classification.The adaptive genetic algorithm is introduced to optimize the wavelet moment,and the algorithm is used to optimize the parameters of the support vector machine.The results show that the optimized data feature aggregation degree increases,and the support vector machine feature classification ability is enhanced.(4)Using MATLAB to analyze the combined moment data,select the combined moment as the input,construct the combined moment,and use the improved support vector machine algorithm to identify the combined moment data for the pattern recognition tool and establish the axis trajectory identification method.Using LabVIEW to design the fault diagnosis system software,the rotating mechanical fault simulation experiment was completed on the rotor test rig,the axial trajectory data was obtained,and the relevant features were extracted.The above identification method was verified.The results show that the fault identification efficiency and accuracy of this method are higher than other features and common classification algorithms.