Research On Fault Detection Method Of Rolling Bearing Fault Based On Fractal

As a key part of rotating machinery,bearing has the advantages of compact structure,high precision,good exchange performance and low frictional resistance.However,its radial size is large,its ability to withstand impact is poor,and it is noisy and has a low service life at high speeds.In the event of a problem,it will lead to a reduction in the life of the equipment,and in the case of an incalculable loss to the country and the people,it is of great significance for the fault diagnosis of rolling bearings.The paper takes rolling bearing as the research object,describe the potential dynamics of the bearing vibration signal from the fractal angle,extract the fault characteristic parameters,and diagnose the fault condition and the degree of fault damage.Pecific research content includes:(1)Research on Bearing Fault Diagnosis Based on Single Fractal.Discuss the influence of noise on the reliability of correlation dimension through simulation experiments and propose the algorithm of correlation dimension based on VMD optimization.Using different fault test bench data to analyze the optimization method,The results show that the correlation dimension can quantitatively describe the dynamic characteristics of the vibration signal,and as the feature vector can effectively distinguish the fault state of the vibration signal and the damage degree of the fault signal;Use the distance function as the index to compare the correlation dimension and the box dimension in the failure mode identification.The results show that the correlation dimension has more advantages in describing the nonlinear complexity of the vibration signal.(2)Constructed a set of characteristic parameters for multifractal detrended wave analysis.The fractal dimension can only describe the dynamic characteristics of the signal as a whole,and it is impossible to meticulate the partial features of the signal.the multi-fractal detrended wave analysis method can make up for the deficiency of fractal dimension,and the multi-scale reveals the dynamic structure of the vibration signal;The multifractal detrended wave method is studied to analyze the multifractal characteristics of different fault signals and different damage degree signals.Based on the fault type and fault size,there are obvious differences in the vibration signal size,and the multi-fractal feature parameters of the vibration signal are extracted to form a high-dimensional feature vector.(3)Research on Bearing Fault Diagnosis Based on Random Forest Compared with the commonly used classification algorithms,random forest has the characteristics of strong applicability,high classification accuracy and strong generalization ability.Therefore,this paper applies random forest algorithm to bearing fault diagnosis.The random forest algorithm voting rule will weaken the shortcomings of the decision tree with high classification accuracy,and propose a random forest algorithm based on weighted improvement.Through experimental verification,the improved random forest algorithm has obvious advantages in bearing fault diagnosis compared with the traditional algorithm.(4)A bearing fault diagnosis system based on fractal theory is designed and implemented.At thend of this paper,the bearing fault diagnosis system based on the above theory is realized.Firstly,according to the demand analysis,the overall architecture design of the system is carried out.Secondly,the function is designed for the database.Finally,the basic functions of each module of the system are debugged to ensure the integrity of the system and the feasibility of the previous research theory.