Research On Dynamic Threshold Calculation Method For Bearing Of Helicopter Swashplate

The bearing is the key component of the helicopter swashplate,and it is easy to occur fault of corrosion,distortion,scuffing and so on during the working process.It will bring significant security threats to the stability and safety of the helicopter flight if the faults were not timely detected.In order to protect the flight safety of the helicopter,real-time monitoring of the working state of helicopter dynamic components,especially swashplate bearings,is one of the important contents of Health and Usage Monitoring System(HUMS).Therefore,it is of great practical significance to study the dynamic threshold calculation method for swashplate bearing.In this thesis,the dynamic threshold calculation method for helicopter swashplate bearing is carried out by using Bayesian inference method and Receiver Operating Characteristic(ROC)curve method.The main work content and research results are as follows:(1)The background,significance and related theoretical basis of the dynamic threshold calculation method for helicopter swashplate bearing are introduced.Firstly,the research background is described in detail,and the classical data processing and dynamic threshold calculation methods are summarized.Secondly,the theoretical basis of vibration signal preprocessing and feature extraction methods are described in detail.(2)A dynamic threshold calculation method for helicopter swashplate bearing based on Bayesian inference is proposed.Firstly,the singular value decomposition is used to preprocess the vibration signal,and the influence of noise on the dynamic threshold calculation result is reduced.Secondly,the fault frequency energy and M6 A feature extracting methods are used to extract the features of normal data and inner ring,outer ring and ball fault data.Finally,Bayesian inference method is used to calculate the dynamic threshold of normal data and inner ring,outer ring and ball fault data respectively.The dynamic threshold experiments are carried out by using vibration data collected by Luoyang LYC Bearing Co.,Ltd.The results show that compare with the weighted average method,the proposed method which has higher true positive rate and lower false positive rate can correctly find the dividing point between the probability density curve of normal data and different fault type data.The proposed method effectively improves the calculation accuracy of dynamic threshold.(3)A dynamic threshold calculation method for helicopter swashplate bearing based on ROC curve is proposed.Firstly,the singular value decomposition is used to preprocess the vibration signal,and the influence of noise on the dynamic thresholdcalculation result is reduced.Secondly,the fault frequency energy and M6 A feature extracting methods are used to extract the features of normal data and inner ring,outer ring and ball fault data.Finally,ROC curve method is used to calculate the dynamic threshold of normal data and inner ring,outer ring and ball fault data respectively.This thesis takes the vibration data collected by Luoyang LYC Bearing Co.,Ltd as experimental data to carry out the dynamic threshold experiments,and the experimental results are compared with the weighted average method and the Bayesian inference method.The experimental results show the results obtained by ROC curve method and Bayesian inference method for calculating the dynamic threshold are very close,and the dynamic threshold can be effectively calculated in the case while most of the original vibration signal is good;In the case while a small part of the original vibration signal is not ideal,there are crossed parts at the left and right ends of the probability density curve of normal data and different fault type data.Therefore,the dynamic threshold calculation results are affected.ROC curve method not only solves the problem,but also improves the calculation accuracy of dynamic threshold when the probability density curve of two types of data have large overlap area.