Research On Fault Diagnosis Method Of Planetary Gearbox Based On Time Series

In the fields of aviation,mining and new energy,the transmission system of rotary machinery is the key to the normal operation of the equipment.The transmission system often works in time-varying conditions.High-speed,variable-load,and limited installation space make planetary gear trains widely used.In these areas.Planetary gearboxes therefore typically operate in harsh environments where gear failure is highly prone to occur.Therefore,it is extremely important for planetary gearbox fault diagnosis to ensure the safety of operators and equipment,and to reduce the cost of equipment maintenance.At present,most of the planetary gearbox fault diagnosis is based on the reciprocating vibration signal for fault analysis.The acceleration sensor installed on the gearbox is used to collect the reciprocating vibration signal required for the test.This method is difficult to extract the gear meshing vibration information.The torsional vibration signal extracted by the incremental encoder installed in the gearbox shafting is much more convenient,because the torsional vibration signal has a relatively simple spectral structure relative to the reciprocating vibration signal,and is not affected by other vibrations,and the fault is also more sensitive.Therefore,this paper proposes a test analysis of the planetary gearbox fault characteristics based on the torsional vibration signal.1.The planetary gearbox torsional vibration signal is collected by the incremental encoder,and the acquired signal is converted into a curve relationship diagram by the principle of high frequency pulse counting,and then the first-order linear interpolation method is used to twist these uneven signals.The vibration signal is converted into a uniform torsional vibration signal.In order to get closer to the planetary gearbox working environment,the magnetic powder brake is used to increase the load on the planetary gear output shaft,and the test overall equipment is perfected,and the torsional vibration signal required for the test is collected.2.The construction and development test platform is used to simulate the actual real working conditions of the planetary gearbox.By calling Labview graphical programming language,the planetary gearbox torsional vibration signal acquisition system needed in the test platform is researched and developed.The computer master control device will store and store the collected torsional vibration signals in txt format for subsequent data processing and analysis.The torsional vibration signal of thecollected planetary gearbox is extracted for fault characteristics,and the time domain and frequency domain of the torsional vibration signal are analyzed,and the time domain and frequency domain diagrams of the normal state of the planetary gear and the fault state are compared.3.The feature extraction of torsional vibration signal is carried out for the working state of the simulated planetary gearbox in the actual environment.The fault diagnosis and prediction of the planetary gearbox based on time series analysis are proposed.The time series model is established.The information theory criterion is used to determine the model order.The improve covariance method is used to estimate the model parameters.When the gear fault state is found,the model coefficient is not within the parameter tolerance range,and the fault signal is predicted.It is found that this method has a good prospect in gear fault diagnosis.