Research On Health Condition Monitoring Method Of Planetary Gear Box

Planetary gearbox is an important structural component of mechanical transmission.It has many advantages in structure,such as light weight,small size,large transmission ratio,strong carrying capacity and high transmission efficiency.It is widely used in aerospace,wind power,shipbuilding and coal field.However,planetary gearboxes usually operate under complicated and severe working conditions such as high temperature,heavy load,corrosion,fatigue,etc.Long-term operation in these complicated and harsh environments will cause fatigue wear on the metal surface of the gears,and the degree of fatigue wear is divided into slight.Moderate and severe,in severe cases,gear breakage will occur,resulting in equipment failure,which has a double impact on economic efficiency and personnel safety.Therefore,monitoring the health status of planetary gearboxes is of great significance.The complexity of the planetary gearbox structure and motion pattern determines the complexity of vibration response and wear changes during operation.At the same time,planetary gearboxes operate in complex and variable environments,causing their loads and speeds to change over time,which will inevitably increase the difficulty of monitoring planetary gearboxes.In response to the above questions,this paper has carried out a series of studies,the main work includes:(1)According to the operating environment of the planetary gearbox,design and build an experimental platform for planetary gearbox condition monitoring and detail the data acquisition system for vibration monitoring and online oil monitoring.According to the expected deterioration degree of the planetary gearbox health status,design and process gears with mild pitting,moderate pitting,severe pitting and broken teeth.Design experimental schemes and set experimental parameters for condition monitoring of planetary gears in different states.(2)This paper introduces the processing method of vibration signal,and uses non-dimensional analysis,envelope spectrum analysis,cepstrum analysis and wavelet analysis to analyze the experimental data of five healthy states of the collected planetary gearbox.By calculating the characteristics of the vibration signals in the five states,the sensitivity of the kurtosis index,the peak index,the pulse index and the margin index are verified,and a multi-parameter diagnostic model is formed.The envelope spectrum in the frequency domain analysis is used to judge and verify the fault location,and the cepstrum is used to realize the simple and rapid diagnosis of the planetary gear health.The application of wavelet analysis in the diagnosis of planetary gears is studied.It is shown that wavelet decomposition can efficiently acquire the characteristic frequency band and intuitively judge the health of different states,The peak-to-peak value of the d3 layer wavelet is extracted as an eigenvalue.Finally,the multi-parameter dimensionless,wavelet decomposition d3 layer peak-to-peak value and the combination of the two characteristics are taken as the eigenvalues respectively.The RVM model is used to realize the health status recognition of the planetary gearbox.By contrast,the two features are combined.The feature quantity is better than the single feature value for the classification of the health status of the planetary gearbox.(3)The online oil-abrasive analysis index-abrasive concentration IPCA is introduced,and the threshold value is divided by the planetary gearbox IPCA data under normal conditions.The state of the planetary gearbox can be divided into normal state,warning state and Abnormal state.The wear trend of the five states was analyzed by IPCA value,and the granule accumulation TIPCA was introduced to establish the GM(1,1)model to predict the trend of the planetary gearbox.The background value selection of gray GM(1,1)model is optimized.The optimized background value is selected as the best background value to be substituted into the GM(1,1)model to complete the optimization of GM(1,1)model.Through the model test,the prediction effect of improved GM(1,1)is better than the GM(1,1)model.Finally,the two monitoring methods are combined to analyze the results of planetary gearboxes in different states,and good monitoring and diagnosis results are obtained.(4)According to the monitoring method and analysis and processing method of the planetary gearbox,the planetary gearbox health status monitoring and evaluation system based on Java language is designed and developed.