Intelligent Fault Diagnosis And Vibration Simulation Of Gear In Planetary Gearbox

Planetary gearbox is one of the most commonly used transmission devices in mechanical system.Its healthy operation is particularly important for the normal work of the system,and it is also an important research object in the field of fault diagnosis.In engin eering practice,planetary gear boxes usually operate in harsh environment and complex working conditions,which always leads to pitting,falling off and cracking of key components.After the local fault s occur,it will further lead to the failure of the whole mechanical system if is working continuously,which may cause heavy property losses and even casualties.Ther efore,it has practical theoretical significance and potential application prospects that to effectively diagnose the local faults of planetary gearbox.The vibration information is an intuitive response to the condition of the equipment.The analysis of t he vibration signal can effectively determine whether there is a fault in the gearbox.As the latest research achievement in the field of pattern recognition and machine learning,deep learning has achieved rich achievements in the fields of image and spee ch processing,and has also made some progress in intelligent fault diagnosis of planetary gearbox.However,there are still many problems to be solved.On the other hand,the verification of the new intelligent fault diagnosis method needs a lot of reliable data.The vibration signal simulation model can provide reliable and convenient data for the verification of the new intelligent fault diagnosis method and other technical methods.However,there are some differences between the existing planetary gearb ox vibration signal simulation model and the actual vibration signal,which will have a certain impact on the verification of the new method.Therefore,it is necessary to study the simulation modeling of planetary gearbox vibration signal which is close to the actual vibration signal.In this paper,the intelligent fault diagnosis method and vibration simulation modeling of typical planetary gearbox are deeply studied.Aiming at the shortcomings of the existing gearbox fault diagnosis methods based on machine learning algorithm,an intelligent fault diagnosis method of planetary gearbox based on long short time memory(LSTM)neural network and feature enhancement is proposed.Firstly,the vibration signals of different local faults of the planetary gearbox are intercepted by sliding windowing.T hen the fast Fourier transform is performed on each segment of the intercepted signals.The frequency band containing rich fault features is selected as the input to train the LSTM neural network.Through the trained neural network model,the fault features in the selected frequency band are intelligently extracted to realize the fault diagnosis of different local faults of the planetary gearbox Identify the diagnosis.In order to solve the problem that the existing vibration simulation model of planetary gearbox can not diagnose the fault effectively by using windowed vibration separation technology and wavelet transform,a vibration simulation model of planetary gearbox based on gear meshing response and meshing seq uence is proposed in this paper.Based on the impact response of gear meshing,the single meshing impact vibration response is simulated firstly,and then the time point of each gear meshing is calculated.The single meshing impact response is spliced acco rding to the gear meshing sequence.The vibration simulation model of planetary gearbox is established considering the time-varying transmission path of vibration signal and the modulation effect of the rotation frequency of sun gear,planetary gear and pl anetary carrier.Through the test on the planetary gearbox transmission test platform,the proposed intelligent fault diagnosis method is verified by using the simulation and measured signals,and the established vibration signal simulation model is compared with the measured signal,which verifies the effectiveness and correctness of the proposed intelligent fault diagnosis method and vibration signal model.