Research On Development And Fault Diagnosis Of Ultra Low Speed Slewing Bearing Test Bench Based On Mechanical Power Closurer

As an important rotating part in major equipment,ultra-low speed slewing bearing is widely used in industrial robots,shield machines,excavators and wind turbines.It has the characteristics of large size,low operating speed(usually less than 10rpm)and large load.Because the working environment of slewing bearing is usually bad,it is one of the most easily damaged parts in rotary machinery and equipment.Therefore,it is particularly important to study its condition monitoring and fault diagnosis and fully grasp its health status.Aiming at the importance of condition monitoring and fault diagnosis of ultra-low speed slewing bearing and the structural characteristics of internal gear slewing bearing,this paper proposes a new distributed mechanical power closed slewing bearing fault diagnosis test bench,and uses multi-sensor information fusion technology to carry out four kinds of health conditions.The research content provides a new technical means for the experimental research and fault diagnosis of ultra-low speed slewing bearing.Firstly,a new type of distributed mechanical power-closed slewing bearing fault diagnosis test bench is proposed.The structural characteristics of the gearbox and loading mechanism in the test bench are explained,and the power-closed principle and actual power loss of the test bench are analyzed.Through dynamic simulation software(Romax),the power-closed transmission chain of the test bench is analyzed,and the results show that the closed power(test power)in the transmission chain of the test bench is not provided by the total power input into the system.The total power input into the system accounts for about 20% of the closed power,only overcoming the power loss caused by friction,stirring oil,and other factors in the system.Therefore,the test bench can achieve energy saving during use.Secondly,the measurement and control system of the test bench is designed.A multimotor synchronous control method based on current signals is studied to realize the synchronous operation of each motor in the driving system of the test bench.The data acquisition system and its software are designed using Lab VIEW.Finally,the energysaving performance test of the complete test bench system is carried out,and the results show that the closed power(test power)in the test bench system is not provided by the driving motor.Through analysis,the power consumed by the driving motor is the loss power caused by overcoming friction in the system,heat generation in the driving system,and power consumption in the reducer.Finally,the multi-sensor information fusion method for slewing bearing fault diagnosis is studied.As one of the classical models of deep learning,convolutional neural network has strong data information mining and information fusion capabilities.Based on the existing MCCNN model,a multi-sensor information fusion method is proposed by adding a channel attention module.The evaluation scores of each channel obtained by the softmax function are applied to the multi-channels of each input layer to improve the feature extraction ability of the original model.Finally,the working condition tests of four health states of slewing bearings,namely normal state,internal gear fracture,internal gear wear,and lubricating oil impurities,are conducted using the developed test bench.The results show that the improved model(MCCNN-CAM)has good fault diagnosis recognition ability.By comparing MCCNN-CAM with single-channel CNN model,it is verified that using multi-sensors for fault diagnosis has higher accuracy,and MCCNNCAM has higher classification accuracy than other multi-sensor information fusion methods such as DS-evidence method.Figure [52] Table [15] Reference [85]...