A Method Of Extracting Characteristic Parameters And Fault Diagnosis For Eliminating Operating Conditions Of Helicopter Main Reducer Input Terminal

Helicopters are important dual-use products for both military and civilian use.The main drive system is the core of the helicopter.Its stability is directly related to the safety of the helicopter.The main reducer is the key to the main drive system.Therefore,online monitoring and fault diagnosis of the main reducer input terminal is of vital importance to guarantee the normal flight of the helicopter.In the actual flight process of the helicopter,affected by the flight environment,the vibration signals collected are generally the vibration signals of variable working conditions.Compared with the vibration signals under steady working conditions,the processing of the vibration signals of variable working conditions and the extraction of fault information are more important.difficult.The purpose of this paper is to find a suitable method to process the vibration signal of the variable working condition,extract the fault characteristics from it,and then combine the method of pattern recognition to perform the intelligent fault diagnosis of the working condition of the helicopter main reducer input terminal.The main research contents of the thesis are:(1)Research the method of removing the working condition of the main reducer input terminal based on Computed Order Tracking(COT).Under the premise that the speed is known,the method of order tracking is used to transform the non-stationary signal in the time domain into a stable signal in the angle domain;when the speed is unknown,the LMS system needs to first extract the rotation frequency from the original signal,and then use the COT.After obtaining the angle-domain stationary signal,perform order spectrum analysis on it.Simulation and experimental signals verify the feasibility of COT to eliminate working conditions.(2)Propose a method based on the Improved Nuisance Attribute Projection(INAP)to eliminate the working conditions of the main reducer input terminal features.Aiming at the shortcoming that the COT method can only remove the speed,the main reducer input terminal features removal method based on the Nuisance Attribute Projection(NAP)is proposed.It is found that the method of constructing the weight matrix by the NAP method cannot quantify the degree of interference of different working conditions to the vibration signal,which will affect the effect of removing the working conditions.Aiming at the shortcomings of the NAP algorithm,INAP is proposed.Firstly,singular value decomposition(SVD)is used to obtain the eigenvalues of the covariance matrix of the eigenmatrix,and then the discrete degree of the eigenvalues is used to quantify the redundant information contained in the vibration signal under a certain working condition.The INAP algorithm is applied to the fault diagnosis of the main reducer input terminal under variable working conditions,and the feasibility of the INAP method to eliminate the working conditions and the advantages compared with the NAP method are verified through simulation and experiment.(3)Propose a fault diagnosis method based on INAP and Kernel Nearest Neighbor Convex Hull Classification(KNNCHC)for helicopter main reducer input terminal working condition fault diagnosis.This method is used in the single fault diagnosis experiment and compound fault diagnosis experiment of the simulation experiment platform.Analysis results verify the advantages of KNNCHC method and the fault diagnosis method based on INAP and KNNCHC can diagnose single fault and compound fault of bevel gearbox under variable working conditions.