Study On Bearing Fault Detection Methods Of Vibrating Screen Based On Enhanced Energy Operators

In recent decades,the vibrating machinery has been rapidly developed and widely used in a large number of production fields,such as mining,metallurgy,and building,etc.Vibrating machinery plays an important player in industrial production.Thus,it is critical for vibrating machinery to maintain a healthy working condition.Rolling element bearing is one of the most important parts and frequently-used elements in vibrating screen.Once the bearings are damaged,it will strongly affect the normal work of the vibrating screen,resulting in the reduction of screening accuracy.Thus,bearing fault detection and diagnosis is very significant to prevent the occasional catastrophic,with serious consequences for safety,production loss,and repair cost.Based on the limitation and expandability of the traditional energy operator theory,this dissertation presents several new on-line diagnosis methods for the unique characteristics of bearing fault signals extracted from vibration screen.The main work of this dissertation is as follows:Therefore,the main purpose of this dissertation is to take the online monitoring as a premise and a vibrating screen as a typical case to study the bearing fault dynamic model and signal characteristics of vibrating machinery.According to the characteristics of online monitoring and the signals extracted from vibrating screen,several novel online monitoring techniques based on the theory of the energy operators are proposed in order to overcome the shortcomings of the existing energy operators.These new techniques adopt new energy transformation and do not need envelope analysis like the traditional energy operators,so the alternative operators with new energy transformation are more robust than the traditional energy operators.The validity and practicability of the theory and algorithm are verified by simulation and real test rig.The main work of the thesis is as follows:1.Considering that the shortcomings of neglecting slip in previous bearing failure dynamic model of vibrating screen,the bearing fault dynamic model of the vibrating screen is established based on the differential motion equation of five degrees-freedom.The difference between the bearing fault signals of vibrating machinery and rotating machinery is analyzed in detail by the model.2.In view that the shortcomings of conventional energy operators and by studying the high-order differential energy operators and analytic energy operators,it is found that the two kinds of energy operators are robust to vibration interferences and background noise,respectively.Thus,a high order differential analytic energy operator(HOD_AEO)based on the two methods is proposed.This kind of energy operator has the advantages of the first two kinds of energy operators and can extract the bearing fault characteristic frequency under the strong noise and vibration interferences.3.Considering that the symmetric difference sequence belongs to the first order derivation and can play the role of smooth filtering,a symmetric difference analytic energy operator(SD_AEO)is proposed.Like the HOD_AEO technique,the SD_AEO can also detect the weak bearing fault signature in the presence of strong noise and vibration interferences.4.A frequency weighted analytic energy operator(FW_AEO)is proposed,whose energy expression greatly differs from those of the first two energy operators.However,like HOD_AEO and SD_AEO,this approach is also capable of identifying the bearing fault characteristic frequency from a heavily contaminated signal.5.According to the characteristics of Multi-faults feature extraction,an enhanced method for compound faults based on a combination of Multipoint Optimal Minimum Entropy Deconvolution(MOMEDA)and Sparse Bayesian step-filtering(SBSF)is proposed.First,MOMEDA can highlight the impulsive components featuring the inner and outer fault characteristics,respectively.And then,SBSF can further remove a lot of background noise.Finally,the enhanced energy operator is utilized to distinguish the inner-and outer fault characteristic frequencies.The effectiveness of this method is verified by experimental analysis.6.In order to better apply the three enhanced energy operators to different working conditions,the applicable scope of the three enhanced energy operators is analyzed.The results of simulation and bearing fault detection experiments indicate that these proposed methods can effectively extract fault features,certifying their feasibility and superiority,and the purpose of on-line monitoring for bearing fault diagnosis of vibrating screen is realized.