| Helicopters have been widely used in the field of national economy and national defense construction,such as geological exploration,emergency rescue,fire rescue,public security patrol and so on.It plays an important role in the development of national economy and national security.As an important part of the helicopter,the working environment of the transmission system is poor and its own structure is complex,which makes the transmission system is prone to failure..Once the transmission system failure,it often lead to serious accidents.The structure of helicopter transmission system is complex,which makes the fault data by experiment is difficult to obtain.To solve this problem,this paper through the method of fault modeling,establishing helicopter transmission system with broken tooth fault,pitting fault and crack fault,and then through the simulation to analyze the evolution law of fault gear transmission system.A new method of gear fault diagnosis based on Gauss mixture hidden Markov model(GMM-HMM)is proposed.The main contents of this papers are as follows:(1)Gear Modeling and Simulation Analysis of Helicopter Transmission System.The paper,uses the main reducer of a model helicopter as the object of study.,Using UG to establish three-dimensional solid model of helicopter gear transmission system,through professional interface software UG and ADAMS,the 3D solid model of gear transmission system in Parasolid format into ADAMS,having flexible processing to ring gear,.Adding motion pair,contact force,drive and load etc.,so as to ensure the smooth operation of the gear transmission system.Calculating the contact stiffness coefficient of the gear and setting the force index,the damping coefficient,the maximum penetration depth,the friction force and so on of the contact force.To determine the most suitable contact stiffness and damping coefficient of the simulation,then conducting the simulation analysis.Utilizing the transmission ratio and meshing frequency to verify the correctness of the model,and verified by experiments,which prove the correctness of the modeling method.(2)Modeling and Simulation Analysis of Typical Gear Fault in Helicopter Transmission System.Based on the established faultless helicopter gear transmission system,the UG characteristic command was used to establish 1/4 broken teeth,2/4 broken teeth,3/4 broken teeth,all broken teeth,two kinds of pitting faults and crack fault gears,And then in the Parasolid format to export,and finally according to the normal sun wheel position and direction coordinates,importing the fault gear into the helicopter gear transmission system,replacing the normal gear,establishing teeth,pitting or crack failure helicopter gear transmission system respectively.Analyzing the influence of the broken teeth on the contact stiffness and adding the constraint pair,the contact force,the driving,and the load.Calculating the contact stiffness coefficient and then analyzed.The time domain,frequencydomain,contact force and delta potential energy of the gear system under four kinds of fault conditions are compared and analyzed.It is found that the amplitude of the time domain,frequency domain,contact force and delta potential energy In the growing,indicating that with the deepening of the fault,the impact between the gears is increasing.It is found that the amplitude of the time domain,frequency domain,contact force and delta potential energy In the growing.Analyzing the time domain and frequency domain of the vibration signal of the helicopter transmission system with pitting or crack failure,and the correctness of the gear modeling and the correctness of the gear fault implantation method are verified.Finally,it is verified for experiment.(3)According to the vibration acceleration signal obtained by ADAMS simulation,the Gaussian mixed hidden Markov model is trained by using four kinds of acceleration signals,such as normal,broken tooth,pitting and crack as sample input.Because the gear vibration signal has different frequency domain capability,The wavelet packet decomposition and reconstruction are carried out on the gear fault signal,and the wavelet packet coefficients are obtained and the wavelet packet coefficients are reconstructed.According to the difference of the energy distribution of the gear vibration signal in the frequency domain under different fault conditions,First,reconstructing the wavelet coefficients to extract the feature vector,and then normalizing the feature vector,of the wavelet packet decomposition and reconstructed fault signals to improve the resolution and achieve the purpose of denoising.Establishing the Gaussian mixture hidden Markov model library,and then re-collecting the signals in these four states and inputting into the Gaussian mixture hidden Markov model which was established with the simulation data,estimating the diagnosis effect based on simulation method on hidden Markov model. |
PDF Full Text Request