Research On Fault Diagnosis Technology Of Rolling Bearing Under Different Speed Conditions

As a key component widely used in mechanical equipment,the health status of rolling bearing directly affects the running performance of mechanical equipment.Affected by various factors,rolling bearing is prone to failure.Therefore,carrying out the necessary bearing maintenance is the key to ensuring the normal operation of mechanical equipment.However,the monitoring signal of the bearing is mixed with a large amount of noise interference,and it is difficult for the traditional fault diagnosis technology to effectively identify the fault of the bearing.In addition,the characteristics of fault signals generated by bearings under different rotational speed conditions are also different.The traditional bearing fault diagnosis technology under constant rotational speed conditions is difficult to be directly applied to bearing fault diagnosis under rotational speed changes.Therefore,in view of the different fault signal characteristics of rolling bearings under different rotational speed conditions,this thesis studies the rolling bearing fault diagnosis methods under two conditions of constant rotational speed and variable rotational speed.The main research contents are as follows:1.An experimental platform for rolling bearing fault signal acquisition is built.Carrying out the bearing fault signal acquisition experiment,and collecting the vibration signals of the bearing under the condition of constant speed and variable speed respectively,to provide data support for the research on bearing fault diagnosis technology.2.Aiming at the problem that it is difficult for traditional fault diagnosis algorithms to effectively extract bearing fault features from complex signals,a fault diagnosis method for rolling bearings based on dictionary learning and spectral averaging is proposed.In this method,the signal is first segmented in the time domain.Then a multi-scale alternating direction multiplier method for dictionary learning proposed in this thesis is used to extract the fault impulse signal from the segmented signal.Finally,the amplitude of the characteristic frequency spectrum of the bearing fault in the envelope spectrum is enhanced by the spectral averaging method,the fault location of the bearing is effectively identified,and the fault diagnosis of the rolling bearing under the condition of constant speed is realized.3.Aiming at the problem that the traditional envelope analysis is difficult to effectively identify the characteristic frequency of aperiodic bearing faults under variable speed conditions,a tacho-less order tracking method based on generalized demodulation is proposed.The method firstly uses the generalized demodulation algorithm to directly extract the rotational speed information of the rolling bearing from the vibration signal.Then the angular domain resampling algorithm is used to convert the signal into a periodic signal in the angular domain according to the extracted rotational speed information.Finally,the fault of the rolling bearing is identified by the order envelope spectrum,and the fault diagnosis of the rolling bearing under the condition of variable speed is realized.4.An online monitoring and fault diagnosis system for rolling bearings is designed and developed.First,selecting the corresponding sensors,signal acquisition instruments and other hardware equipment to build a data acquisition system;then,the fault diagnosis algorithm of rolling bearing proposed in this thesis is applied to the system,and the system software module is designed.Finally,the human-computer interaction interface of the system is designed to facilitate users to operate the system.